<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Syahrier Wakid on Medium]]></title>
        <description><![CDATA[Stories by Syahrier Wakid on Medium]]></description>
        <link>https://medium.com/@syahrier-wakid?source=rss-b4fd4860eafd------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*rtWGpRgmM3qmIOahNkXrqw.jpeg</url>
            <title>Stories by Syahrier Wakid on Medium</title>
            <link>https://medium.com/@syahrier-wakid?source=rss-b4fd4860eafd------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 18:55:13 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@syahrier-wakid/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Redefining Customer Experience with Hybrid Intelligence]]></title>
            <link>https://syahrier-wakid.medium.com/redefining-customer-experience-with-hybrid-intelligence-767dc6a4dcc0?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/767dc6a4dcc0</guid>
            <category><![CDATA[customer-experience]]></category>
            <category><![CDATA[hybrid-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-and-humans]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Tue, 30 Nov 2021 07:35:00 GMT</pubDate>
            <atom:updated>2021-11-30T15:15:54.193Z</atom:updated>
            <content:encoded><![CDATA[<h4>“Brings The Best of Machine and Human Intelligence Together To Elevate Customer Experience For Brand Across The Globe”</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/0*uxAuw5QUqfiSJm8o" /><figcaption>Image source:whyy.org</figcaption></figure><blockquote><strong>2020 was the year no one of us expected.</strong></blockquote><p>The COVID-19 pandemic hit all of us and instantly pushed companies all over the world to transform their business. COVID-19 changed the way we connect with our friends, family, colleagues as well as changed how customers and companies communicate to each other. Companies need to work extra hard to find new creative digital solutions out of necessity to keep them connected with their customer in real-time despite offline store closed, social distancing and work from home policy. Boundaries between our online and offline lives dissolved. <strong>These trends of communication</strong> <strong>are here to stay even after the pandemic</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/583/0*F6rtIhYGChvT062n" /><figcaption>Image source: 105.net</figcaption></figure><p>Twilio surveyed over 2,500 businesses globally to better understand the effect of COVID-19 on the company’s digital transformation and communication roadmap. The result is that, COVID-19 accelerated companies’ digital communications strategy by six years (Twilio Report, 2021). In brief, a decade-long digital transformation roadmaps of almost every business all over the globe, got compressed into weeks even days in order to adapt fast because of the COVID-19 effect. Businesses in every industry had to figure out how to reach their customers seamlessly, real-time and across channel.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rJZ1wUaapzvdqYmDMSSXng.png" /><figcaption>Data source: Twilio COVID-19 Report</figcaption></figure><p><strong>So, what can we do as a business owner to ensure customer engagement through COVID-19 and even beyond?</strong></p><p>Before answering that question, let’s take a look at another important study from NICE inContact (Nasdaq: NICE) found that Millennials and Generation Z (let’s call them <strong>Generation Now</strong>) are the main force of digital-first future of Customer Experience as they forms the largest part of the global consumer base (Businesswire, 2019). This is also true in Indonesia as the number of Generation Z reaches 75.49 million and the millennial generation reaches 69.90 million (Kurniawan, 2021). Almost half of Indonesia’s population are <strong>Generation Now</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Fck3DyqGRYvDwJzc" /><figcaption>Image credit: NeuAPIX-Yellow.ai</figcaption></figure><p>Another finding from NICE inContact is that almost 60% of Generation Now have used private social messaging to contact business representatives or customer services. They also want businesses to allow them to interact using social messaging apps (72% and 69%, respectively). It is understandable since the generation now is digital savvy. They were born in the era of smartphones and other high-tech devices with Artificial Intelligence (Siri, Alexa, Cortana, Google Assistant. In result, they have a much shorter attention span, are impatient, they are extremely comfortable doing multitasking and expect personalized engagement, more importantly, using social media as their main tools to contact business representatives (Alexandra).</p><p>Now, we can answer our questions above, the future solutions that business owners must prepare to compete in the era of pandemic and beyond as well as with the digital savvy consumers is redefining how businesses interact with consumers with the implementation of AI and omni-channels platforms including private social messaging are fundamental to deliver seamless, real-time and exceptional customer experiences across channel.</p><p><strong>How exactly is the exceptional CX that exceeds the expectations of the digital-savvy customers? The answer is Hybrid Intelligence.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/786/0*Lmf6wFIfUI_v1J-f" /><figcaption>Image source: McKinsey &amp; Company</figcaption></figure><p><strong>Hybrid Intelligence</strong> is a collaborative intelligence between humans and AI that are joining forces (Wilson &amp; Dougherty, 2019)<strong>. </strong>But, many of us will think why not just implement only an AI for our business? Not to mention that we see how AI is depicted in the movies such as Iron Man with its super AI, Jarvis that can understand complex order and can think by itself. Well, up until today, nobody knows how to build Jarvis-like AI. Whether this is even possible in the future is still up to debate.</p><p><strong>The current advanced technology of AI is that it is becoming good at many </strong>“human” jobs, especially the repetitive and simple ones such as data entry, bookkeeping, providing customer service virtual assistant (voicebot and chatbot), automatically reserve restaurant, ticketing system, hotel booking, even translating language and manufacturing assembly. But, there are still lots of shortcomings that AI cannot yet perform complex tasks alone, such as applying knowledge to decision making or creative tasks.</p><p>In particular, current AI technology still has significant difficulties adapting to ever-changing environments and self-adjusting, they lack what we all call common sense (Lake et al., 2017). AI is exceptional at conducting specific, well-defined tasks based on a particular type of data and in a controlled environment, AI is also good at finding complex and subtle patterns in huge sets of data. Not to mention that they don’t get bored and deliver speed and accuracy (Wilson &amp; Dougherty, 2019).</p><blockquote>​​”Humans and machines can enhance each other’s strengths.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZP9ixwgTFfEn3L2-wXPbsg.png" /><figcaption>Image source: hrb.org</figcaption></figure><p>On the contrary, humans outperformed machines in creative tasks, complex problems that need complex decision making and dynamic social interaction tasks. Not to mention that humans can learn from only a few examples (while AI needs to be trained with a massive amount of datasets) and work efficiently even when there is noise and uncertainty (unpredictable environments). This is why humans and AI need to work together and join forces as <strong>hybrid intelligence</strong>.</p><blockquote><strong>The Future of Customer Experience with Hybrid Intelligence</strong></blockquote><p>Businesses must innovate their services into hybrid intelligence and handle issues, complaints, booking, payments and other digital services seamlessly, real-time and personalized every customer engagement with speed and accuracy while maintaining the human touch. Let’s take a look at one of the example below on how the hybrid intelligence works:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GqwAhyjXtikuzgO9" /><figcaption>Image source: NeuAPIX-Yellow.ai</figcaption></figure><p>In the picture above, AI become the center but do not forget that behind the screen, live agent orchestrate everything, from preparing the knowledge base to answer the query that the AI cannot respond ‘yet’ meaning that it needs to be train again and input the result into the knowledge base so in the future when the similar question being asked again, now the bot can answer it.</p><p>This process shows that AI amplifies our cognitive strengths, interacting with customers and employees with accuracy to free the live agent to do higher-level tasks while still embodying human capabilities. With AI delivering services with an all-important empathetic “human touch”, the role of live agents is becoming the key in the future of Customer Experience.</p><blockquote>“Human and AI, preserves the “human touch” that personalizes every customer interaction with speed and accuracy.”</blockquote><p>Another important reason why hybrid intelligence is important to deliver exceptional Customer Experience in the pandemic era and beyond is that businesses must avoid making <strong>complicated customer journeys</strong>. The most common pain point are the brand owner customer service not always available at 24/7, brand owner cannot follow all the queries that come from customers in different channels (SMS, WhatsApp, Facebook, Instagram, Telegram, Twitter) all at once, customers’ issues hardly resolved in one interaction and forced to repeat themselves each time they communicate to the customer service, brand owner customer service get repetitive questions and transactions or use cases.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*zbn6raCwxHYfCAph" /><figcaption>Image source: NeuAPIX-Yellow.ai</figcaption></figure><p>We can see from the picture above that initially customers had difficulty contacting the brand owner due to limited channels, long processing time, not completed in one contact and channels that were not integrated. Now, customers have easy access to brand owners quickly, seamlessly and integrated this all due to the collaborative intelligence between humans and machines.</p><p><strong>Conclusion</strong></p><p>Avoiding the impact of the AI revolution in the workplace is no longer an option. Everyone across the organization, be it fresh-level employees or senior managers, management and executives as well as C-level, should embrace AI as an essential tool that will improve business productivity, employee well-being and recruitment processes. We will be working alongside AI, organizations that use AI only to replace humans through the automation process will miss the full potential of AI capabilities. Such strategy is erroneous from the get-go.</p><blockquote>The future of Customer Experience is bright, when we have the right capabilities to be prepared for a world of <strong>Hybrid Intelligence</strong>.</blockquote><p>Tomorrow’s leaders will instead be those that embrace collaborative intelligence, transforming their operations, their markets, their industries, and — no less important — their workforces. — <a href="https://hbr.org/search?term=h.%20james%20wilson&amp;search_type=search-all"><strong>H. James Wilson</strong></a><strong>*</strong></p><p>*Accenture Global Managing Director of Information Technology and Business Research</p><p><strong>References:</strong></p><p>Alexandra. CQ Net Online. The future of work is Hybrid Intelligence: How Artificial Intelligence and Human Intelligence support each other. <a href="https://www.ckju.net/en/blog/future-work-hybrid-intelligence-how-artificial-intelligence-and-human-intelligence-support-each-other/59482">https://www.ckju.net/en/blog/future-work-hybrid-intelligence-how-artificial-intelligence-and-human-intelligence-support-each-other/59482</a></p><p>Businesswire. (2019). Millennials and Generation Z Are Driving the Digital-first Future of Customer Experience. <a href="https://www.bloomberg.com/press-releases/2019-12-03/millennials-and-generation-z-are-driving-the-digital-first-future-of-customer-experience">https://www.bloomberg.com/press-releases/2019-12-03/millennials-and-generation-z-are-driving-the-digital-first-future-of-customer-experience</a></p><p>Kurniawan, D<strong>. </strong>(2021). VOI Online. BPS: Indonesia’s Population Is Dominated By Humans Born In 1997–2012 And 1981–1996<strong>. </strong><a href="https://voi.id/en/news/28251/bps-indonesias-population-is-dominated-by-humans-born-in-1997-2012-and-1981-1996">https://voi.id/en/news/28251/bps-indonesias-population-is-dominated-by-humans-born-in-1997-2012-and-1981-1996</a></p><p>Lake B. M., Ullman T. D., Tenenbaum J. B., and Gershman S. J. (2017). <em>Building Machines that Learn and Think Like People</em>. Behavioral and Brain Sciences, vol. 40.</p><p>Twilio Report. (2021). <em>Twilio Covid 19 Digital Engagement Report. </em><a href="https://www.twilio.com/covid-19-digital-engagement-report">https://www.twilio.com/covid-19-digital-engagement-report</a></p><p>Wilson, J &amp; Daugherty, P. (2018). Collaborative Intelligence: Humans and AI Are Joining Forces. From the Magazine (July–August 2018). <a href="https://hbr.org/2018/07/collaborative-intelligence-humans-and-ai-are-joining-forces">https://hbr.org/2018/07/collaborative-intelligence-humans-and-ai-are-joining-forces</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=767dc6a4dcc0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Rise of Generation Mute: The End of Phone Calls Era]]></title>
            <link>https://syahrier-wakid.medium.com/the-rise-of-generation-mute-the-end-of-phone-calls-era-112f6b88eaa2?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/112f6b88eaa2</guid>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[generation-mute]]></category>
            <category><![CDATA[phone-call]]></category>
            <category><![CDATA[omnichannel]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Fri, 16 Apr 2021 09:20:41 GMT</pubDate>
            <atom:updated>2021-04-16T09:20:41.070Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*83IJMCsH83JxiLcLmH78Gw.jpeg" /><figcaption>Image source: youthemploymentdecade.org</figcaption></figure><p>Cut off contact with the closest people seems impossible in this digital era. This is because there are many digital platforms that make communication easy, both via text (chat), video calls (video calls), and phone calls (phone calls). However, long before the digital era, we use to have a cable phone in our house and the famous phone booth called “Wartel”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*kSgsF_-K5Hd6xnDLPbKr8A.jpeg" /><figcaption>image source: <a href="https://merahputih.com/post/read/wartel-yang-punah-karena-teknologi-digital">https://merahputih.com</a></figcaption></figure><p>The generation born in the 2000s may not be familiar with the “magic booth” called wartel. Wartel which stands for “Warung Telekomunikasi” or Telecommunication Shop was popular in the 90s, when cellphones were still expensive things that not everyone could afford. Wartel provide landlines for anyone who wants to contact family, friends, or loved ones far away.</p><p>Wartel was at its peak in 1990–2000. At that time, Wartel business owners had become a source of prosperous income. This is because people need telephone communication so they will look for telephone shops. No wonder so many people are willing to queue for a long time to wait for their turn to call.</p><p>According to data from the Ministry of Communication and Informatics (Menkominfo) of the Republic of Indonesia, the existence of Wartel has started to decline in 2010. In that year the number of Wartel connections was recorded at 282,412. The most significant decrease occurred in 2014, where telephone connections became 836 from the previous year, which was 227,55. Sadly, in 2015 there were no more Wartel lines in all parts of Indonesia.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/750/1*u7V3SROyMKryZzt0BwW_8g.png" /><figcaption>Image sourece: katadata</figcaption></figure><p>The lesson from the loss of Wartel is that they are not prepare to anticipate the advanced development of technology, and as a consequence the Wartel now is just become a “history”. Technology has shifting very fast into the next level in line with the advent of cell phones and smartphones.</p><p>Not only that, the emergence of text-based applications, from the heyday of Mig33 to today’s WhatsApp, Line, WeChat, Telegram as well as social media chatting features (Facebook Messenger, Instagram, Twitter) significantly reduce the use of phone lines.</p><p>Moreover, recent studies shows that millennials — born in the 1980s to 1997 — have been noted to avoid phone calls more often. The preferred form of communication is through text or even more extreme: innuendo through Instagram Story updates. At least so according to a survey from gadget retailer “BankMyCell”.</p><blockquote>“We arrive at a condition where the millennial generation has another name: <strong>The Generation Mute</strong>. Which mean that millennial prefers text-based communication rather than a phone call.”</blockquote><p>The ‘generation mute’ trend is all too apparent, people growing up in the digital age have adopted alternate forms of communication causing a decline in voice calls. Surfacing technologies like SMS, MSN Messenger, BBM, and Myspace have all paved the way for today’s most popular forms of digital communication such as WhatsApp, Facebook Messenger, and Snapchat. (Source: <a href="https://www.bankmycell.com/blog/why-millennials-ignore-calls">https://www.bankmycell.com/blog/why-millennials-ignore-calls</a>)</p><p>Here are the data from BankMyCell regarding the reasons why Millenials do not want to pick a phone call:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/842/1*KMvjuunvd4LukmxjE26ecg.png" /></figure><p>Based on the survey above, wouldn’t we rather have just sent an text message or used a messaging app? No wonder Generation Mute hates phone calls. They’re a horribly inefficient way of conducting business, especially now that digital options are available.</p><p>Another unique finding, 81 percent of millennial respondents in the study said they were often nervous when they wanted to call. This tendency is termed <strong>“Telephonophobia”</strong>. Therefore, one in five respondents admitted that they had to prepare themselves before starting to do a conversation through telephone. The conclusion of the survey teaches this, because millennials generally grow in front of the screen.</p><p>Let’s see another survey from MarketerMag regarding the phone calls which shows that there’s nothing so special about phone calls anymore for either the customer or the business. Here is the list why millennial do not want to use a phone call based communication:</p><ol><li><strong>Calling a customer service is annoying.</strong> When’s the last time you looked forward to having to call an Customer Service number and navigate a lengthy, confusing automated menu before getting to talk to an agent?</li><li><strong>Calling a customer service is time-consuming.</strong> It takes to long to navigate that automated menu, then sit on hold waiting for a reply, then have the agent put us on hold at various times throughout the call to look up information or update your account. In between work and other urgent activities, who has time for this?</li><li><strong>Calling a customer service is inefficient and ineffective.</strong> The representative must keep looking up things about your account, updating various parts of the system, and possibly querying supervisors. Meanwhile, you must explain your situation or issue from scratch each and every time you call. This is certainly not effective.</li><li><strong>Customer service usually not operate effectively for 24/7</strong>. You’ll probably have to call sometime during the workday if you want good service. It is means you will most likely have to call while you’re on the bus or train, at work, or in some other public place.</li><li><strong>Customer service call center not making it private conversation</strong>. No matter how low you keep your voice, someone is bound to hear, and maybe you don’t want everyone in the world knowing how you spend your money and how much you spend.</li></ol><p>Contacting millennials by voice call nowadays is becoming increasingly difficult, particularly for marketing purpose. Businesses should understand why people in the digital era are avoiding calls and look to evolve their outreach with another communications technologies that suit this generation better.</p><blockquote>“Ditching customer service hotlines and focusing modern digital channels won’t just elevate your brand in the eyes of Generation Mute. It will also enable your organization to offer far superior service to customers across all generations and at a lower cost… Now isn’t that something! “ <a href="https://marketermag.com/generation-mutes-aversion-to-phone-calls-is-upending-customer-service/">https://marketermag.com/generation-mutes-aversion-to-phone-calls-is-upending-customer-service/</a></blockquote><p>Let’s see a great video about Generation mute from BBC United Kingdom.</p><p>&lt;iframe src=”https://www.bbc.co.uk/ideas/videos/generation-mute-no-talking-aloud/p05ydw78/player&quot; width=”500&quot; height=”440&quot; scrolling=”no” style=”overflow: hidden” allowfullscreen frameborder=”0&quot;&gt;&lt;/iframe&gt;</p><p><strong>Conclusion:</strong></p><p>Nowadays, our smartphones are so advanced and able of making communications that don’t require voice calls, that make voice calls aren’t as important anymore. Companies that want to connect with these costumers had better do so through messaging channels (chatbot) or leveraging voice calls with Artificial Intelligence to make it into voicebot.</p><p>In brief<strong>, </strong>preparing advanced messaging channels communication could help the company stay ahead in today’s fierce competition and follow the customers characteristics and set out with their evolving preferences now and make real-time support available both through a call and through a instant messaging (omnichannel platform).</p><blockquote><em>“The biggest part of our digital transformation is changing the way we think.”<br></em><strong><em> — </em></strong><a href="https://www.inspiringalley.com/digital-transformation-quotes/"><strong>Simeon Preston</strong></a><strong>, Bupa</strong></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=112f6b88eaa2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Membuat Sendiri Virtual Assistant Seperti Jarvis Milik Iron Man Tanpa Coding]]></title>
            <link>https://syahrier-wakid.medium.com/membuat-sendiri-virtual-assistant-seperti-jarvis-milik-iron-man-tanpa-coding-47ef23db6910?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/47ef23db6910</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[google-dialogflow]]></category>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[virtual-assistant]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Sun, 26 Jul 2020 07:55:16 GMT</pubDate>
            <atom:updated>2020-07-26T07:55:16.560Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SwoYU6Lpt0_pY_czyFgEig.jpeg" /><figcaption>Image source : Phillip Tracy novatiosolutions.com</figcaption></figure><p><strong>Just A Rather Very Intelligent System</strong> atau yang lebih dikenal dengan J.A.R.V.I.S adalah sistem kecerdasan buatan yang diciptakan oleh si Iron Man, Tony Stark.</p><p><strong>Trivia:</strong></p><blockquote>Dalam komiknya, nama J.A.R.V.I.S sebenarnya diambil dari nama Edwin Jarvis, <em>butler </em>(pelayan) yang sebelumnya bekerja untuk Howard Stark, ayah dari si Tony Stark.</blockquote><blockquote>Seiring berjalannya waktu, Tony Stark mengembangkan Jarvis menjadi sistem kecerdasan buatan sekaligus menjadi asisten super Tony Stark yang mengurus seluruh keperluannya, termasuk menjadi “Otak” dari <em>Armor </em>tempur Iron Man.</blockquote><p>Saat kita menonton Iron Man, pasti banyak yang bertanya-tanya pada diri sendiri, betapa asyiknya jika bisa memiliki J.A.R.V.I.S.</p><p>Bayangkan betapa kerennya jika kita memiliki asisten AI di rumah, misal saat pulang ke rumah kemudian disambut oleh si asisten AI, “Welcome home, Sir”. sembari kita bilang, “JARVIS, turn on the light. please. Super cool isn’t it?</p><p>Sebenarnya AI assistant ini sudah ada di dunia nyata, beberapa contohnya adalah, Google Assistant yang berada di hampir semua gadget Android, atau SIRI untuk yang mempunyai gadget iPhone. Ada juga Cortana, yang ada di perangkat Microsoft, Alexa, virtual assistant yang dibuat oleh Amazon. Kemudian yang paling baru yaitu Bixby, untuk perangkat electronic Samsung.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OMQjSyNFF_PE63Ms-sF6tA.png" /></figure><p>Nah, kabar baiknya, kita bisa membuat sendiri AI Assistant <strong>tanpa coding.</strong> Wait, tanpa coding? yes, Google telah menyediakan platform untuk membuat sendiri AI Assistant, tentunya tidak akan secanggih J.A.R.V.I.S.</p><p>Sebenarnya untuk saat ini, <strong>belum ada</strong> teknologi pada bidang <em>Artificial Intelligence </em>yang bisa membuat <em>self-consious </em>atau dalam artian kecerdasan buatan yang bisa berpikir sendiri dan membuat keputusan sendiri seperti layaknya manusia. Tetapi, saat ini juga para ilmuwan sedang berlomba-lomba untuk membuatnya, sehingga mungkin saja dalam waktu dekat akan ada asisten pribadi sepintar J.A.R.V.I.S.</p><p>Oke, let’s stop the blabbering and get to the tutorial :D</p><p><strong>Google Dialogflow</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*qh9GsinDGeB0guSS4dVu2w.png" /><figcaption>Image Source: Google</figcaption></figure><blockquote>Dialogflow is a natural language understanding platform used to design and integrate a conversational user interface into mobile apps, web applications, devices, bots, interactive voice response systems.</blockquote><p>Secara umum, Dialogflow adalah framework milik Google yang menyediakan servis Natural Language Processing (NLP) dan Natural Language Understanding (NLU) yang umum digunakan untuk membuat sebuah chatbot. Lebih dari itu, Dialogflow memiliki fitur <strong>one-click</strong>, yaitu dapat terintegrasi dengan mudah ke sebagian besar platform populer seperti Facebook, Telegram, Twitter, Viber, Kik, dll. Dan juga dapat terintegrasi dengan Google Assistant dan Amazon Alexa.</p><p><strong>Berikut langkah-langkah membuat sendiri asisten virtual-mu di Google Dialogflow:</strong></p><ol><li><strong>Membuat Akun Google Mail</strong></li></ol><p>Kunjungi <a href="https://mail.google.com/">https://mail.google.com</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/805/1*9kPRNEg5d5twOGl2b9myWg.png" /><figcaption>Image source: Google</figcaption></figure><p><strong>2. Membuat Akses Dialogflow</strong></p><p>Kunjungi <a href="https://dialogflow.cloud.google.com/">https://dialogflow.cloud.google.com</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R7bXwVExvcfKv0Bj6-h0Ug.png" /><figcaption>Image source: Google</figcaption></figure><p><strong>3. Membuat Agent (Virtual Assistant)</strong></p><p>Setelah membuat Akun Gmail and akses ke Dialogflow, langkah berikutya adahal membuat <strong>Agent</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vkudUXROoNUz2XzLm9WMNA.png" /></figure><p>Agent yang dimaksud di dalam platform ini adalah <strong>robot/bot</strong> yang nantinya dapat berperan sebagai <strong>Virtual Assistant</strong> kita. Agent dicontoh ini bernama <strong>“Manis”. </strong>Di sini kita bisa memilih bahasa yang diinginkan, untuk kali ini saya memilih menggunakan Bahasa Inggris. Dan untuk zona waktu, saya memilih Asia/Bangkok.</p><p><strong>4. Membuat Intent dari Agent</strong></p><p>Apa yang dimaksud dengan <strong>intent</strong>? Di sini intent adalah tujuan dari percapakapan, misal dalam contoh ini, kita akan membuat si Manis menyapa balik ketika kita mengetik “Hi”. “Hello”, “Hello, Manis”, “Hi, Manis”.</p><p>Dan kata sapaan di atas masuk sebagai intent untuk Manis<strong>. </strong>Secara Default, Dialogflow memiliki 2 intent default yaitu <strong>Default Welcome Intent </strong>dan <strong>Default FallBack Intent.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hl7CGFdqVpT9TAwIMzYxDA.png" /></figure><p><strong>5. Membuat Default Welcom Intent</strong></p><p><strong>Default Welcome Intent </strong>disini adalah kalimat yang akan disampaikan ketika kita menyapa Manis untuk pertama kali nya, contohnya seperti ini. Pertama kita input kata-kata sapaan tadi ke dalam <strong>Training Phrase.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ns65vRw8YYjFXEk2KM78YQ.png" /></figure><p>Kemudian, kita juga harus mempersiapkan respond dari Manis ketika dia mendapatkan sapaan.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/775/1*TZuAhuF3h5CgWArflynZ-w.png" /></figure><p><strong>6. Membuat Default Fallbcak Intent</strong></p><p><strong>Default Fallback Intent</strong> adalah pesan yang akan disampaikan jika kalimat yang disampaikan oleh kita sebagai pengguna tidak terkait dengan intent yang sebelumnya kita buat di Default Welcome Intent, misal kita menyapa dengan kata <strong>“Hungry”. </strong>Karena kalimat sapaan tersebut tidak ada dalam kata sapaan maka kita harus menyiapkan kalimat untuk agent. Contohnya<strong> “I didn’t get that. Can you say it again, Sir?”</strong>,<strong> “I missed what you said. What was that, Sir?”. </strong>Dengan kata lain Fallback intent ini adalah semacam pesan error.</p><p><strong>Mari kita coba kemampuan si Manis.</strong></p><p>Cara mencobanya, cukup ketik kata sapaan di input form di sebelah</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/364/1*OjT2GXfJEbB-HS2vVMG-6A.png" /></figure><p>Terlihat bahwa, Manis mengerti dengan sapaan yang kita berikan. Kita coba dengan kalimat yang belum kita latih.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/344/1*DLznI-AReaxexbRWto9-vw.png" /></figure><p>Yang keluar adalah <strong>Default Fallback Intent</strong>. Karena agent belum dilatih untuk menjawab pertanyaan diluar daftar sapaan pada <strong>Default Welcome Intent</strong> yang dipersiapkan sebelumnya.</p><p><strong>Mencoba Kemampuan Agent dengan Web Demo Text Based:</strong></p><p>Selain bisa mencoba kemampuan agent langsung dengan dialog form, kita juga dapat mencoba kemampuan agent dengan cara mengintegrasikannya beragam. Cara mengintegrasikannya, klik Integration di sebelah kiri website:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/238/1*H1ZcAMew5Pwgu3Qv_hnFrg.png" /></figure><p>Lalu halaman akan berpindah ke pilihan integrasi yang diinginkan, seperti gambar di bawahi ini:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/709/1*Xk3cWBxwVZ2SBZL134Of9g.png" /><figcaption>Text based chat integration</figcaption></figure><p>Untuk contoh dalam tulisan ini, kita akan menggunakan Web Demo. Klik bagian web demo, lalu akan muncul pop-up message seperti di bawah ini:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kvSs3Zd9ixm-by0LtTKNyw.png" /></figure><p>Klik tombol <strong>“Switch” </strong>pada pojok kanan atas, makan pop-up message akan berubah seperti di bawah ini.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V5MdfInPxpzUhvs2KYsRqA.png" /></figure><p>Kemudian, klik link yang ada di dalam pop-up message dan halaman akan berpindah seperti di bawah ini.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9AXe9kKwDzAMXscnakZ41Q.png" /><figcaption>Text based web demo Dialogflow</figcaption></figure><p>Terlihat bahwa berhasil merespon percapakan dengan baik dan sesuai yang diinginkan. Untuk mempelajari detail tentang dialogflow, berikut adalah akses beberapa konsep dasar Dialogflow dengan mempelajari dokumentasi padahl tautan ini <a href="https://cloud.google.com/dialogflow/docs/">https://cloud.google.com/dialogflow/docs/</a></p><p><strong>Kesimpulan</strong></p><p>Tulisan bagian pertama ini sebagai gambaran langkah awal membuat virtual assistant dan bagaimana membuat interaksi secara sederhana dengan assistant tersebut. Di bagian selanjutnya saya akan menunjukkan kepada bagaimana melatih dan membuat percakapan yang lebih rumit serta bagaimana mengintegrasikan dengan platform lain.</p><p><strong>Stay tune!</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=47ef23db6910" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Breaking Prejudice with Logic: 12 Angry Men Movie Review and Analysis]]></title>
            <link>https://syahrier-wakid.medium.com/breaking-prejudice-with-logic-12-angry-men-movie-review-and-analysis-6f74c20b14d8?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/6f74c20b14d8</guid>
            <category><![CDATA[prejudice]]></category>
            <category><![CDATA[movies]]></category>
            <category><![CDATA[logic]]></category>
            <category><![CDATA[12-angry-men]]></category>
            <category><![CDATA[movie-review]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Tue, 31 Mar 2020 22:43:20 GMT</pubDate>
            <atom:updated>2020-04-01T05:26:14.147Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*diWCHp7szgRNdXhgjtoi4w.jpeg" /><figcaption>Image source: Everett Collection.</figcaption></figure><p><em>Prejudice is a learned trait. You’re not born prejudiced; you’re taught it. — </em><a href="https://www.brainyquote.com/authors/charles-r-swindoll-quotes"><em>Charles R. Swindoll</em></a></p><p>Have you ever thought or assumed something negative about a group and its members? When you visit an area, have you ever assumed that the people there are rude when talking?</p><p>This is a form of prejudice. Prejudice can be found in the community even in yourself when looking at certain groups. You might even have been a victim of that prejudice. So what is prejudice?</p><blockquote>“Prejudice is a negative attitude that one has based mostly on opinions and stereotypes rather than facts/evidence.<em> </em>The word prejudice is of Latin origin, <strong>pre </strong>meaning “before” and prejudice stemming from “judged.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*GV7k0ii9rngyqi500mi_dQ.jpeg" /><figcaption>Image source: BBC</figcaption></figure><p>Talking about prejudice, there is a movie that really show how a prejudice can influence heavily on people’s assessment towards something. Without further ado, let’s review this one of the best movies.</p><p>The film starts in a courtroom. Inside the courtroom, there was a trial against a child who was accused of murder of his own father. If the allegation is proven, then the child will be sentenced to death. Regarding whether or not the allegations are proven, it all depends on the 12 jurors who have seen the proceedings of the trial from the beginning to the end. The common law court system adopted by the United States is a juror who decides whether a defendant is guilty or innocent.</p><p>So, after the prosecutors and defense attorneys completed their duties, the witnesses had testified, and there was enough evidence, the judge gave the 12 jurors, which consist of people from different background, the mandate to negotiate and decided unanimously whether the defendant was found guilty or innocent.</p><p>It is predicted that the decision will be decided shortly due to the evidence, witnesses were all pointing out towards the kids as guilty. That is why, they went directly through the voting mechanism to map the vote. However, the result is beyond prediction, of the 12 jurors, it was ‘only’ 11 people who judged that the defendant was guilty, the one that vote not guilty is juror #8. According to the rule, 11 votes cannot be categorized as acclamation. Therefore, they must decide to round the 11 votes to 12, or vice versa. But it was not an easy task, especially for Juror #8 which is the only one standing for not guilty against 11 other who pressed him because of his choice.</p><p>Juror #8 himself was not sure if the defendant was truly innocent, for him it is not fitting to determine the life of a person in just five minutes. Not that he is convinced the defendant is innocent, but he is only less certain that the defendant is guilty which in the movie called <strong>reasonable doubt</strong>. The other jurors cannot accept this reason. In the stale and hot atmosphere, with the cross of opinion, the war of words, the pounding of various ego, they begin to start debating and arguing to find the correct decision between guilty or not guilty. To make it interesting, almost in the entire movie, they call each other by Juror #1, Juror #2, Juror #3 and so on. But in the end of the movie, we know the name of Juror #8 (Davis) our protagonist and Juror #9 (McCardle).</p><p>In this movie, intelligent dialogues become the main factor. The jurors displayed that they are not just debating without purpose, but they used constructive argument, using facts based on the evidences and witnesses, as well as theories to support or refute the opinion of the other person. Each character is given time and a portion to show off, so that of the 12 jurors characters will have a spotlight. None of the 12 jurors in the movie seem to be redundant, they have their own role to play.</p><p>By watching this movie, we can take some lessons that we can apply in real world situations. The first is the importance of a discussion. It was not a correct way to use a voting, because it will depend on the majority. As reflected in the fourth precepts of Pancasila and the Preamble of the 1945 Constitution of my country, the Republic of Indonesia. Deliberation to consensus is the best way to make a decision, especially if the decision is so important. Because, in this way we can hear each other opinion before making a decision.</p><p>Another lesson is that the majority is not always true and the minority is not necessarily wrong. Minority votes are entitled to be heard not to be directly suppressed by the majority. Juror #8 gives an idea of how effective persuasive communication, which is an effective way to change the minds and behaviors of those with whom you disagree. Even though he becomes a minority, it does not matter. As long as we can back-up our argument.</p><p>In this movie, we also understand that there are different people with different character. Juror #1 who organize thing and lead others, Juror #2 that’s so easy to intimidate, Juror #3 that are so prideful and won’t admit that he is wrong until the last, Juror #4 that are overconfident, Juror #5 that are afraid to state his opinions and had left on a slump, Juror #6 that doesn’t quite understand what is going on, Juror #7 that takes things easily, even if it is a human life, Juror #8 which is our protagonist calm and brilliant, Juror #9 the old man who is so wise, Juror #10 who always angry and let his prejudice clouded his observation, Juror #11 who is a watchmaker and very polite and Juror #12 who just going with the flow.</p><p>Juror #8 also taught us to stay calm, observing other behavior and preparing solid proof. In the end, with a critical and logical and rational analysis supported by the facts, Juror #8 sought to gradually erode the analytical flaws of the witnesses and evidences. Moreover, Juror #8 resolve and determination successfully makes the other Juror change their minds.</p><p>The next lesson is that Juror #8 wants to show us that, if we do not know the truth with certainty, we must think carefully and all personal issues, emotions, and prejudices must be kept away before making decision, search for reasonable doubt. And more importantly, we must admit that if we are wrong, not being stubborn and keep ignoring the fact that presented to us.</p><p>The last and the most important thing, deciding whether a person deserves to die or not is one of the big decisions that we have to take responsibility for the rest of our life, and not a trivial decision that we can decide in a hurry and emotional way.</p><p>In Brief, this movie deserved a 9/10 and credit to Reginald Rose for its fantastic script as well as Sidney Lumet, one of the best movie directors. With limited source, but so impactful, the dialogue, interaction between each character, the argument and reasoning. All of it teach us that we can create something so impactful with only a few resources. No wonder that this movie is labelled as one of the best movies of all time!</p><p>It doesn’t even make sense to judge a person before you know them, it’s unfair to them. Let alone blame someone without proof and clear logic</p><blockquote><em>“Some people are prejudiced because they don’t want to understand another person’s difference, or they feel scared or threatened. “</em></blockquote><p><strong><em>“Do not be ignorance, stand up to prejudice, use your logic and think clearly before you judge someone, you might inspire other people to do the same!”</em></strong></p><p><strong>References:</strong></p><ul><li><a href="https://www.betterhelp.com/advice/psychologists/what-is-prejudice-psychology-definition-and-examples/">What Is Prejudice? Psychology, Definition, And Examples | Betterhelp</a></li><li><a href="https://www.verywellmind.com/what-is-prejudice-2795476">The Psychology Behind People&#39;s Prejudices</a></li><li><a href="https://www.simplypsychology.org/prejudice.html">Prejudice and Discrimination in Psychology | Simply Psychology</a></li><li><a href="https://www.imdb.com/title/tt0050083/?ref_=nv_sr_srsg_0">12 Angry Men (1957) - IMDb</a></li><li><a href="https://www.bbc.co.uk/cbbc/findoutmore/help-me-out-prejudice">Lifebabble - Help me out - prejudice</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6f74c20b14d8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Face Recognition Implementation using Python with Open Source Computer Vision Library (OpenCV)]]></title>
            <link>https://medium.com/analytics-vidhya/face-recognition-implementation-using-python-with-open-source-computer-vision-library-opencv-8454c0aab926?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/8454c0aab926</guid>
            <category><![CDATA[face-recognition]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[image-processing]]></category>
            <category><![CDATA[opencv-python]]></category>
            <category><![CDATA[image-recognition]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Wed, 16 Oct 2019 01:20:38 GMT</pubDate>
            <atom:updated>2020-08-26T05:39:52.257Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/896/1*88ukwmilVBLspWHRqNniaw.png" /><figcaption>Image Source The Sun</figcaption></figure><blockquote><strong>What is Face Recognition?</strong></blockquote><p>Face Recognition is a new category of an Artificial Intelligence (AI) that can map a person’s facial features mathematically and save their data as a face-print. This software uses an algorithmic system to compare direct or digital image capture which is then stored on a face-print. This will be useful for verifying someone’s identity. Basically, face recognition is the same as fingerprint matching technology, retinal scanning, and speech recognition that is done to distinguish between one physical and another. All of these systems take data from strangers, analyze the data in the input, then match it to the entries in the database.</p><p>Face recognition technology is not something foreign in this era. This technology has been applied almost everywhere. For example, Apple, Samsung and Xiaomi have all installed face recognition technology in their smartphone, mostly they used to unlock the phone. Another example, is Facebook automatic tagging feature. Facebook automatically recognize when when people should be tagged in photos. There are other face recognition implementation, facilitate secure transaction, ATM face identifying, airport checking face recognition and many more.</p><blockquote><strong>How Face Does Recognition work?</strong></blockquote><p>According to the article from <strong><em>The Conversation</em></strong>, the face detection process can be taken in three stages: detection, face-print creation, and verification or identification. When an image is captured, the software in the computer will analyze it to identify where the faces are. After a face has been identified, a facial recognition system will process images more closely and then poured in the form of a face-print. Like fingerprints, face-print is a characteristic that is used to specifically identify a person’s face. But, if we want to expand it into details, there are several steps to detect a face. This is the process of the facial recognition system.</p><ol><li><strong>Detection</strong></li></ol><p>Taking photos of human faces is done by scanning 2D photos digitally or using video to take 3D face photos.</p><p><strong>2. Alignment</strong></p><p>When successfully detecting a face, the software will determine the position, size and attitude of the head. The position of the head must face the camera at least 35 degrees and up to 90 degrees.</p><p><strong>3. Measurement</strong></p><p>The image of the captured face then scaled and rotated so that it can be stored and mapped into an appropriate pose and size in the database.</p><p><strong>4. Normalization</strong></p><p>The system reads the geometry of the face by determining their key signature, for example the distance between the eyes, the shape of the nose, the thickness of the lips, the distance between the chin and the forehead, and many others. Noted that the more key signature a system can get, the more powerful the system to recognize a face.</p><p><strong>5. Representation</strong></p><p>After forming the facial signature, the system converts it into a unique code. This coding facilitates easier computational comparison of the newly acquired facial data to stored databases of previously recorded facial data.</p><p><strong>4. Matching</strong></p><p>If the photos on the faces that have been represented and the availability of face photos in the database are both match, the recognition process can be done immediately.</p><blockquote><strong>Introducting OpenCV (Open Source Computer Vision Library)</strong></blockquote><p>Computer Vision is the ability of a machine / computer to see to be able to extract information from an image. One of the fields related to Computer Vision is Image Processing or commonly called Image Processing. OpenCV is a library (library) that is used to process images and videos so that we are able to extract information in it. The word Open in OpenCV is meant to be open source ie free, not paid, can be downloaded by anyone.</p><p>While the CV in the word OpenCV is short for Computer Vision, it means that the computer is used to process images captured by a recording device (such as a camera or webcam) which convert the data from analog to digital and then processed in a computer with lines of code. OpenCV can run in various programming languages, such as C, C ++, Java, Python, and also support on various platforms such as Windows, Linux, Mac OS, iOS and Android.</p><blockquote><strong>Introducting Python</strong></blockquote><p>Python is a multipurpose interpretive programming language that uses a design philosophy with a focus on the level of code readability. As a programming language, Python combines the capabilities, capabilities and syntax of code and high-quality library functions.</p><p>Basically, the programming language designed by Guido van Rossum is actually very widely used to create programs that are very often used by the general public. Call it the GUI program (desktop), smartphone applications, CLI programs, IoT, games, web, hacking programs and much more.</p><p>Python’s ability to process images is very limited, for that he needs to import libraries from OpenCV. This is where the link between OpenCV and Python.</p><blockquote><strong>The Face Recognition Implementation using Python-OpenCV</strong></blockquote><p>To be able to create face recognition using Python and OpenCV. The first step is to install the OpenCV library in Python using pip install:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/428/1*s8H_gNfD2DHakAYOjZ13kg.png" /></figure><p>For this step, we will try to implement the face detection first. Here is the code using Python OpenCV:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/647/1*Kq9RmLEz121b7IPkuxFvCg.png" /></figure><p>To be able to used the code, we have to write this line in the command prompt:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/903/1*GykVu1IhsT1eKefvPJRzfQ.png" /></figure><p><strong>The result:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iLSzudlTQyXzC8itSLDQfg.png" /><figcaption>Image source ABC News</figcaption></figure><p>Another example with eyes detection:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*vC0mYdSNSNAYrN7MEN2ztw.png" /></figure><p><strong>Result:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DqkvHLfvkAcTjS3xhE64_A.png" /></figure><p><em>*The eyes detection is not perfect, especially in the Robert Downey Jr faces since he wear a glasses, also Chris Hemsworth only detect one eyes since he is looking elsewhere.</em></p><p><strong>Notes:</strong></p><p>The code above is Python OpenCV basic implementation for the face recognition which is the face detection. The code successfully detect a face in a photo. Notice that those two code have xml files. It is actually <strong>Cascade Classifier (Haar-Like Feature). </strong>The code detection will not work without the xml files, so you have to download it first from here:</p><p><a href="https://github.com/rk45825243/Face-eye-detection-using-Haar-Cascade-classifier">rk45825243/Face-eye-detection-using-Haar-Cascade-classifier</a></p><p>OpenCV uses a type of face detector called the Haar-cascade classifier. Haar like Feature is a method commonly used in object detection. The name Haar itself refers to Haar Wavelet, a mathematical function that is box-shaped and has principles like the Fourier function. Haar-like features are rectangular features, which give specific indications on an image or image.</p><p>The principle of Haar-like features is to recognize an object based on the simple value of the feature but not the pixel value of the object’s image. This method has the advantage that the computation is very fast, because it only depends on the number of pixels in a square not every pixel value of an image</p><p>The method proposed by Viola and Jones combines four main keys to detect an object (Viola, Paul, &amp; Jones, 2001):</p><ol><li>Simple square feature, called Haar feature</li><li>Integral image for fast feature detection</li><li>The AdaBoost machine-learning method</li><li>Cascade classifier to combine many features</li></ol><p>Haar-like feature processes images in boxes, where in one box there are several pixels. Per the box was then processed and obtained a difference in value (threshold) which indicates the dark and bright areas. These values will be used as the basis for image processing.</p><p><strong><em>*In the next post, I will write in detail about Haar Feature and continue to expand the code so it will be able to recognize face of some people not just detecting the faces.</em></strong></p><p><strong>References</strong></p><p>Viola, Paul, &amp; Jones, M. (2001). Rapid object detection using boosted cascade of simple features. In: Proceedings IEEE Conf. on Computer Vision and Pattern Recognition.</p><p><a href="https://www.facefirst.com/blog/amazing-uses-for-face-recognition-facial-recognition-use-cases/">https://www.facefirst.com/blog/amazing-uses-for-face-recognition-facial-recognition-use-cases/</a></p><p><a href="https://theconversation.com/facial-recognition-is-increasingly-common-but-how-does-it-work-61354">https://theconversation.com/facial-recognition-is-increasingly-common-but-how-does-it-work-61354</a></p><p><a href="https://www.scienceabc.com/innovation/facial-recognition-works.html">https://www.scienceabc.com/innovation/facial-recognition-works.html</a></p><p><a href="https://opencv.org/">https://opencv.org/</a></p><p><a href="https://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_tutorials.html">https://opencv-python-tutroals.readthedocs.io/en/latest/py_tutorials/py_tutorials.html</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8454c0aab926" width="1" height="1" alt=""><hr><p><a href="https://medium.com/analytics-vidhya/face-recognition-implementation-using-python-with-open-source-computer-vision-library-opencv-8454c0aab926">Face Recognition Implementation using Python with Open Source Computer Vision Library (OpenCV)</a> was originally published in <a href="https://medium.com/analytics-vidhya">Analytics Vidhya</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kuliah di Hongaria dengan Beasiswa Stipendium Hungaricum]]></title>
            <link>https://syahrier-wakid.medium.com/kuliah-di-hongaria-dengan-beasiswa-stipendium-hungaricum-78a637ed14ef?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/78a637ed14ef</guid>
            <category><![CDATA[stipendium-hungaricum]]></category>
            <category><![CDATA[hongaria]]></category>
            <category><![CDATA[study-in-hungary]]></category>
            <category><![CDATA[kuliah-di-hongaria]]></category>
            <category><![CDATA[beasiswa]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Sat, 28 Sep 2019 00:02:10 GMT</pubDate>
            <atom:updated>2020-12-06T05:38:47.839Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PgqYSAqO55jlmuP8qH40vA.png" /></figure><p><strong>Tentang Beasiswa Stipendium Hungaricum</strong></p><blockquote>Beasiswa program internasional, perkuliahan menggunakan bahasa Inggris.</blockquote><p>Program Beasiswa Stipendium Hungaricum diluncurkan pada 2013 oleh Pemerintah Hongaria. Program ini dikelola oleh <em>Tempus Public Foundation </em>yang merupakan organisasi <em>non-profit</em> yang didirikan pada tahun 1996 oleh Pemerintah Hongaria.</p><p>Beasiswa Stipendium Hungaricum bertujuan untuk mendukung internasionalisasi pendidikan tinggi Hongaria, memperkuat hubungan internasional komunitas akademik dan penelitian di institusi pendidikan Hongaria, meningkatkan keanekaragaman budaya dan untuk mempromosikan reputasi yang baik dan daya saing pendidikan tinggi Hongaria ke seluruh dunia.</p><p>Program beasiswa <strong>Stipendium Hungaricum</strong> didasarkan pada perjanjian kerja sama pendidikan bilateral yang ditandatangani antara institusi yang bertanggung jawab untuk pendidikan di negara tersebut dengan Hongaria. Indonesia sendiri diwakili oleh Kementrian Pendidikan Dan Kebudayaan sebagai mitra (<em>s</em>e<em>nding partner</em>) untuk Hongaria terkait dengan beasiswa Stipendium Hungaricum.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/426/1*n6WT9dBsgp0RZoVSXUfsig.png" /><figcaption><a href="https://dikti.kemdikbud.go.id/pengumuman/pengumuman-pendaftaran-stipendium-hungaricum-scholarship-programme-tahun-2021-2022/">https://dikti.kemdikbud.go.id/pengumuman/pengumuman-pendaftaran-stipendium-hungaricum-scholarship-programme-tahun-2021-2022/</a></figcaption></figure><blockquote>Sejauh ini selama 2016–2019, Indonesia mendapatkan jatah beasiswa 50 (<strong>update per tahun 2020</strong>, jatah beasiswa untuk Indonesia menjadi 100) orang tiap tahun untuk tingkat pendidikan Undergraduate(S1), Master (S2) dan Doctoral (S3). Untuk jatah per tingkat pendidikan juga bervariasi tiap tahun, tetapi sejauh ini lebih banyak mahasiswa master dibanding doctoral. Apakah dapat berubah? tentu saja, tergantung dari perjanjian antara Indonesia dan Hongaria.</blockquote><p><strong>Timeline Pendaftaran Beasiswa Stipendium Hungaricum</strong></p><p><em>*Timeline pendaftaran berdasarkan kondisi tahun 2018/2019 lalu dan akan diperbarui jika timeline terbaru telah keluar.</em></p><ol><li>15 November*: Sistem pendaftaran online dibuka.</li><li>Januari : Batas waktu pengiriman semua dokumen ke <em>Tempus Public Foundation</em> dan <em>sending partner</em>.</li><li>Maret : Tahap pertama seleksi: Nominasi</li><li>April — Mei : Tahap kedua, prosedur administrasi dari institusi</li><li>Awal Juni : Pemberitahuan hasil seleksi</li><li>Juni — Juli : Proses pembuatan visa</li><li>September : Kedatangan di Hongaria</li></ol><p>*Tentative, bisa saja lebih cepat (Akhir Oktober) atau malah lebih lama (Awal Januari)</p><p><strong>Cara Pendataran Beasiswa Stipendium Hungaricum</strong></p><p>*Sistem pendaftaran Beasiswa ini adalah satu pintu. Jadi, aplikan nanti akan mendaftar beasiswa sekaligus mendaftar kampus tujuan dalam satu website.</p><ol><li><strong>Mengisi Online application form</strong></li></ol><p>Semua aplikan beasiswa Stipendium Hungaricum diharuskan mendaftar terlebih dahulu kemudian mengisi form aplikasi secara online di situs resmi beasiswa yaitu :</p><p><a href="https://apply.stipendiumhungaricum.hu/news/new/35-new-call-applications-fall">Apply online! - Stipendium Hungaricum</a></p><p>Tautan di atas juga dapat digunakan untuk melihat daftar kampus yang termasuk dalam program beasiswa ini. Tidak semua kampus ada program beasiswa ini. Tapi secara umum hampir semua kampus (milik pemerintah) dan beberapa kampus swasta bekerja sama dengan program beasiswa ini.</p><p><strong><em>Note: </em></strong><em>Perlu diingat bahwa aplikan diharuskan mengisi dengan cermat dan mengunggah semua dokumen yang diperlukan di dalam form pendaftaran.</em></p><p><strong>2. Dokumen yang dibutuhkan</strong></p><p>Berikut adalah dokumen dan kelengkapan yang harus sudah disiapkan sebelum mendaftar.</p><ul><li><strong>Scan Ijazah dan transkip nilai dalam Bahasa Inggris.</strong></li></ul><p>Harap disiapkan sejak awal sebelum pendaftaran dibuka untuk alih bahasa ijasah dan transkrip yang masih berbahasa Indonesia. Wajib menggunakan Ijazah dengan alih bahasa Inggris bertanda tangan dan stempel resmi yang diakui pihak kampus asal mahasiswa di Indonesia.</p><ul><li><strong>Motivation Letter</strong></li></ul><p>Minimal 1 halaman dengan ukuran kertas A4, untuk ketentuan font size 12pt, font type Times New Roman, Spasi 1.5. Tetapi dapat berubah sewaktu-waktu, jadi tetap harus dicek ulang dihalaman pendaftaran nanti.</p><ul><li><strong>Scan Passport Terbaru</strong></li></ul><p>Dicek terlebih dahulu masa berlaku passport, apakah masih valid sampai keberangkatan nanti di bulan September.</p><ul><li><strong>Scan Sertifikat Kemampuan Bahasa Inggris</strong></li></ul><p>Untuk sertifikan bahasa Inggris, menggunakan TOEFL iBT dan IETLS. Nilainya pun juga bervariasi tergantung dari jurusan dan tingkat pendidikan (Master atau Doctoral). Tetapi dari pengalaman sebelumnya, untuk Master minimal IELTS 6 dan TOEFL iBT 80.</p><p>Apakah <strong>TOEFL ITP</strong> bisa digunakan? <strong>bisa</strong>, tapi tergantung kebijakan kampus masing-masing, ada yang membolehkan ada yang tidak, <strong>kebanyakan tidak membolehkan</strong>. Jadi untuk lebih amannya, pastikan sudah mempunyai IELTS atau TOEFL iBT atau pastikan dengan pihak kampus.</p><ul><li><strong>Sertifikat Kesehatan bebas HIV/AIDS dan HEPATITIS</strong></li></ul><p>Beasiswa Stipendium Hungaricum mewajibkan para aplikan untuk membuat surat bukti atau sertifikat bebas HIV/AIDS dan Hepatitis A, B, C. Sertifikat ini dapat diminta di rumah sakit terdekat di kota masing-masing aplikan.</p><ul><li><strong>Rencana Riset (optional untuk S2, WAJIB untuk S3)</strong></li></ul><p>Rencana riset juga dibutuhkan terutama untuk para aplikan yang melamar di tingkat pendidikan Doctoral (S3). Untuk syarat dan ketentuannya, tergantung dari kebijakan kampus pilihan masing-masing.</p><ul><li><strong>Surat Rekomendasi (optional untuk S2, wajib untuk S3)</strong></li></ul><p>Surat rekomedasi wajib untuk aplikan yang melamar untuk tingkat pendidikan Doctoral (S3), untuk syarat dan ketentuan pemberi rekomendasi, bergantung pada kebijakan kampus pilihan masing-masing</p><ul><li><strong>Statement of Application</strong></li></ul><p>Dapat diunduh ketika mengakses form pendaftaran.</p><p>* <em>Semua pelamar program doktor diwajibkan untuk mengunjungi </em><a href="http://www.doktori.hu/"><em>http://www.doktori.hu</em></a><em> sebelum mereka mengajukan aplikasi mereka. Situs web ini menyediakan informasi dalam bahasa Inggris dan bahasa Hongaria tentang sekolah doktoral Hongaria, bidang penelitian yang tersedia dan topik penelitian serta pengawas siswa Doktoral.</em></p><p><strong>Skema Beasiswa Stipendium Hungaricum</strong></p><ul><li>Bebas Biaya kuliah selama masa studi di Hongaria</li><li>Uang saku bulanan:</li></ul><p>Tingkat Master jumlah uang saku bulanan HUF (Hungarian Forint) 43 700 atau sekitar Rp. 2.185.000 (kurs 1 HUF = 50) kontribusi untuk biaya hidup di Hongaria, sampai penyelesaian studi.</p><p>Tingkat doktor: sesuai dengan undang-undang Hongaria saat ini, jumlah beasiswa bulanan adalah HUF 140 000 untuk fase pertama pendidikan (4 semester) dan HUF 180.000 untuk fase kedua (4 semester) ) — selama 12 bulan setahun, sampai penyelesaian studi.</p><p><strong>Kontribusi akomodasi:</strong></p><p>Asrama gratis atau jika memilih tinggal di luar asrama mendapatkan uang akomodasi sebesar HUF 40 000 per bulan untuk biaya akomodasi selama durasi beasiswa.</p><p>Harap dicatat bahwa jika siswa tidak tinggal di asrama, maka HUF 40 000 / bulan adalah kontribusi untuk biaya sewa itu, dan di kota-kota besar — terutama di ibu kota — kontribusi ini tidak akan cukup, misal di Budapest karena kisaran menyewa tempat tinggal flat atau apartemen kisaran HUF 55 000 hingga 90 000 tentunya bergantung dengan fasilitas, letak, semakin dekat dengan pusat kota semakin mahal.</p><p>Perlu diketahuhi bahwa para penerima beasiswa nantinya diharapkan menambahkan sumber keuangan mereka sendiri untuk menutupi semua biaya hidup di Hongaria terutama untuk sebulan pertama ketika tiba, karena berdasarkan pengalaman sebelumnya, beasiswa baru turun ketika bulan kedua (jangan khawatir, untuk bulan pertama tetap dapat, hanya waktunya diberikan pada bulan berikutnya, jadi dapat double).</p><p>Dan saat awal kedatangan juga perlu mengurus kartu tiket trasportasi bulanan, membeli kelengkapan seperti bantal, selimut (beberapa tidak disediakan di asrama), baju hangat, keperluan mandi, dan beberapa keperluan lainnya.</p><blockquote>Apakah cukup dengan uang beasiswa yang diberikan? Tentu semua bergantung gaya hidup, tetapi secara umum, bisa dibilang sangat cukup. Jadi tidak perlu khawatir akan kelaparan di sana :D.</blockquote><p><strong>Asuransi kesehatan:</strong></p><p>Penerima beasiswa nantinya akan mendapatkan asuransi kesehatan dari pemerintah Hongaria dan dari pihak swasta yang ditunjuk oleh pemerintah di masing-masing kampus.</p><p>Layanan perawatan kesehatan sesuai dengan undang-undang Hongaria yang relevan (UU No. 80 tahun 1997, kartu asuransi kesehatan nasional) dan asuransi kesehatan tambahan hingga HUF 65 000 per tahun per orang.</p><p>Harap diingat bahwa ketentuan ini hanya merupakan kontribusi untuk pemegang Beasiswa.</p><p><em>Tulisan selanjutnya akan membahas tips dan trik seputar bagaimana cara membuat Motivation Letter yang baik.</em></p><p><strong>Referensi:</strong></p><p><a href="http://studyinhungary.hu/">http://studyinhungary.hu/</a></p><p><a href="https://tka.hu/english">https://tka.hu/english</a></p><p><a href="https://docs.google.com/document/d/1RlSY5074sV1nuPXUZtXm0AkcbgJeQdtBEhU-dp3M6nQ/edit">https://docs.google.com/document/d/1RlSY5074sV1nuPXUZtXm0AkcbgJeQdtBEhU-dp3M6nQ/edit</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78a637ed14ef" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pengenalan Deep Learning (Bagian 2)]]></title>
            <link>https://syahrier-wakid.medium.com/pengenalan-deep-learning-bagian-2-53eb8d777281?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/53eb8d777281</guid>
            <category><![CDATA[tensorflow]]></category>
            <category><![CDATA[sentiment-analysis]]></category>
            <category><![CDATA[dnn-classifier]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Sat, 21 Sep 2019 08:55:24 GMT</pubDate>
            <atom:updated>2019-09-21T08:55:24.658Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Analisa Sentimen dengan TensorFlow dari situs ulasan film IMDB</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/689/1*jHzNpL-KagnaHUSHzPTPkA.jpeg" /><figcaption>Image source: KDNuggets</figcaption></figure><blockquote>What is Sentiment Analysis?</blockquote><blockquote><strong>Sentiment analysis</strong> is a type of data mining that measures the inclination of people’s opinions through Natural Language Processing (NLP), computational linguistics and <strong>text analysis</strong>, which are used to extract and <strong>analyze</strong> subjective information from the Web — mostly social media and similar sources.</blockquote><p>Analisa sentimen<em> </em>merupakan salah satu cabang ilmu pad<em>a Computational Linguistics </em>yang bertujuan untuk mengetahui dan meprediksi opini masyarakat mengenai suatu hal, pada umumnya opini di sosial media atau sumber lainnya yang ada di internet (review film, review produk).</p><p>Tujuan utama untuk penulisan <em>Deep Learning</em> pada bagian 2 kali adalah menunjukan bagaimana implementasi TensorFlow untuk analisa sentimen. Berikut adalah langkah-langkah yang diperlukan:</p><ol><li><strong>Menyiapkan Fitur-fitur TensorFlow untuk Analisa Sentimen</strong></li></ol><p>Karena pada tulisan ini hanya dikhusukan untuk implementasi Python pada Windows 10 menggunakan Python 3.7 maka, cara install TensorFlow dan library-library nya menggunakan <em>pip install </em>pada command prompt windows. Berikut adalah implementasinya:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/725/1*b7_WzUBuWNX61XvXvaHlqg.png" /><figcaption>Command Prompt Windows</figcaption></figure><p>Untuk implementasi kali ini dibutuhkan:</p><ol><li><strong>Pandas</strong></li></ol><blockquote><strong>Pandas</strong> is an open source software library, BSD license, written for the <strong>Python</strong> programming language for data manipulation and analysis. In particular, it offers data structures and data analysis operations for manipulating numerical tables and time series.</blockquote><p><strong>Pandas </strong>adalah salah satu fitur gratis pada Python untuk manipulasi dan analisa data. Khususnya pada operasi numerik dan data yang berdasaran waktu.</p><p><a href="https://pandas.pydata.org/">pandas</a></p><p><strong>2. Numpy</strong></p><p><em>NumPy is the fundamental package for scientific computing in Python. It is a Python library that provides a multidimensional array object, various derived objects (such as masked arrays and matrices), and an assortment of routines for fast operations on arrays, including mathematical, logical, shape manipulation, sorting, selecting, I/O, discrete Fourier transforms, basic linear algebra, basic statistical operations, random simulation and much more.</em></p><p><strong>Numpy </strong>adalah fitur gratis pada Python yang memiliki fungis-fungsi dasar aritmatik. Contohnya aljabar linear, statistika, simulasi dan beberapa lainnya.</p><p><a href="https://docs.scipy.org/doc/numpy-1.13.0/user/whatisnumpy.html">What is NumPy? - NumPy v1.13 Manual</a></p><p><strong>3. Movie review datasets</strong></p><p>Pada implementasi kali ini, digunakan <em>datasets movie review </em>dari salah satu situs ulasan film terbesar di internet, IMDB. Dataset terdiri dari 25.000 ulasan film sangat dengan tingkatan tingkatan review (menggunakan bintang) untuk pelatihan (train dataset), dan 25.000 untuk pengujian (test dataset). Kita akan menggunakan dataset ini untuk melatih model klasifikasi biner, yang dapat memprediksi apakah ulasan tersebut positif atau negatif.</p><p><a href="http://ai.stanford.edu/~amaas/data/sentiment/">Sentiment Analysis</a></p><p>Tujuan dari implementasi ini adalah memberi label pada ulasan tersebut sebagai negatif atau positif. Berikut adalah langkah-langkahnya.</p><ol><li><strong>Inisialisasi Libraries Python</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/372/1*fMehaafC9rd2ikRkcga96w.png" /></figure><p><em>*keterangan:</em></p><p><strong>import os</strong></p><pre><em>The OS module in Python provides a way of using operating system dependent functionality. <br><br>The functions that the OS module provides allows you to interface with the underlying operating system that Python is running on – be that Windows, Mac or Linux. You can find important information about your location or about the process.</em></pre><p><strong>import re</strong></p><pre>Python has a built-in package called re, which can be used to work with  Regular Expressions.</pre><p><strong>2. Unduh database melalui Python</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*TIt2nSvQBG-P-Lluwpf2Jg.png" /></figure><p>*keterangan:</p><pre><strong>Keras</strong> is a high-level API to build and train deep learning models. It’s used for fast prototyping, advanced research, and production, with three key advantages:</pre><blockquote><em>User friendly<br>Keras has a simple, consistent interface optimized for common use cases. It provides clear and actionable feedback for user errors.</em></blockquote><blockquote><em>Modular and composable<br>Keras models are made by connecting configurable building blocks together, with few restrictions.</em></blockquote><blockquote><em>Easy to extend<br>Write custom building blocks to express new ideas for research. Create new layers, loss functions, and develop state-of-the-art models.</em></blockquote><p><strong>Keras </strong>merupakan fungsi API (Application Programming Interface) level tinggi pada <em>model deep learning.</em></p><blockquote><strong>tf.keras </strong>is TensorFlow’s implementation of the <a href="https://keras.io/">Keras API specification</a>. This is a high-level API to build and train models that includes first-class support for TensorFlow-specific functionality, such as <a href="https://www.tensorflow.org/guide/keras#eager_execution">eager execution</a>, <a href="https://www.tensorflow.org/api_docs/python/tf/data">tf.data</a> pipelines, and <a href="https://www.tensorflow.org/guide/estimators">Estimators</a>. <a href="https://www.tensorflow.org/api_docs/python/tf/keras">tf.keras</a> makes TensorFlow easier to use without sacrificing flexibility and performance.</blockquote><p><strong>3. Menyiapkan kolom yang dibutuhkan</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/652/1*4utoYldiQBRK8OlVy-n6lw.png" /></figure><p>*keterangan:</p><p>Fungsi <strong>load_directory_data</strong> adalah fungsi untuk memuat semua file dari direktori dalam DataFrame. Dan membuat dua kolom, yaitu kolom sentence dan kolom sentiment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/628/1*S1VbFyNycOj-4LyO18MvQw.png" /></figure><p>*keterangan:</p><p>Fungsi load_dataset ini adalah fungsi untuk menggabungkan contoh ulasan positif dan negatif pada dua folder, dan menambahkannya kedalam kolom dengan nama polarity yang memiliki nilai 1 untuk positif dan 0 untuk negatif. untuk melihat data lengkapnya setelah proses unduh selesai, lakukan perintah berikut:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/606/1*SRJ1Giygo94IPRfWJd2cbA.png" /></figure><p><strong>4. Membuat model Deep Learning</strong></p><p>Menggunakan <strong>tf.estimator. </strong>Estimator adalah representasi tingkat tinggi TensorFlow untuk model yang lengkap. Fitur ini menangani detail inisialisasi, logging, menyimpan dan memulihkan, dan banyak fitur lainnya. Untuk lebih jelasnya, silakan melihat pada tautan berikut:</p><p><a href="https://www.tensorflow.org/guide/premade_estimators#overview_of_programming_with_estimators">Premade Estimators | TensorFlow Core</a></p><p>Implementasi pada Python:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/643/1*kPsdO4uzovxpCFgO3Ge47g.png" /></figure><p><strong>Kolom fitur</strong><br>TF-Hub menyediakan kolom fitur yang menerapkan modul pada fitur teks yang diberikan dan meneruskan output modul. Dalam tutorial ini kita akan menggunakan modul nnlm-en-dim128.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/499/1*Cu-FW05kE7ZR59b_DQme3w.png" /></figure><p>Untuk klasifikasi kita dapat menggunakan <strong>DNN (Deep Neural Network) Classifier, </strong>yaitu<strong> </strong>salah satu dari classifier yang digunakan pada TensorFlow. Perintah ini membangun jaringan saraf multilayer (Multilayer Neural Network) feedforward yang dilatih dengan satu set data berlabel untuk melakukan klasifikasi pada data yang serupa dan tidak berlabel. Untuk lebih jelasnya, dapat dibaca melalui tautan berikut:</p><p><a href="https://www.mapleprimes.com/maplesoftblog/209354-A-Beginners-Guide-To-Using-The-DNN">https://www.mapleprimes.com/maplesoftblog/209354-A-Beginners-Guide-To-Using-The-DNN</a></p><p>Implementasi Python:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/555/1*ykueaYiv9ey1UF1lknAPlg.png" /></figure><p>Jalankan prediksi untuk data pelatihan (train dataset) dan data tes (test dataset).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/615/1*-Ie391apngxPz7OGry0FWg.png" /></figure><p>Untuk tahap selanjutnya adalah improvement pada hasil akurasinya.</p><p><strong>References</strong></p><p><a href="https://www.pythonforbeginners.com/os/pythons-os-module">https://www.pythonforbeginners.com/os/pythons-os-module</a></p><p><a href="https://medium.com/tensorflow/building-a-text-classification-model-with-tensorflow-hub-and-estimators-3169e7aa568">https://medium.com/tensorflow/building-a-text-classification-model-with-tensorflow-hub-and-estimators-3169e7aa568</a></p><p><a href="https://www.tensorflow.org/guide/keras">https://www.tensorflow.org/guide/keras</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=53eb8d777281" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pengenalan Deep Learning (Bagian 1)]]></title>
            <link>https://syahrier-wakid.medium.com/pengenalan-deep-learning-bagian-1-8b84c751ae34?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/8b84c751ae34</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[tensorflow]]></category>
            <category><![CDATA[deep-learning-framework]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Mon, 09 Sep 2019 13:42:02 GMT</pubDate>
            <atom:updated>2019-09-09T13:42:35.980Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>Implementasi menggunakan Deep Learning Framework TensorFlow</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YrvMKrWMhi3HomoiTLPsfw.png" /><figcaption>image source: TensorFlow Github</figcaption></figure><blockquote>What is Deep Learning?</blockquote><p>“Deep learning is an AI function that mimics the workings of the human brain in processing data for use in decision making. Deep learning AI is able to learn from data that is both unstructured and unlabeled. Deep learning is a machine learning technique that teaches computers to do what comes naturally to humans: learn by example. Inspired by the structure and function of the brain called <strong>artificial neural networks</strong>.”</p><p>Jadi, <strong>Deep Learning</strong> adalah fungsi kecerdasan buatan (Artificial Intelligence) yang meniru cara kerja otak manusia dalam memproses data dan menciptakan pola untuk digunakan dalam pengambilan keputusan. Deep Learning juga merupakan cabang ilmu dari <strong>Machine Learning</strong> dalam AI yang memiliki jaringan yang mampu belajar tanpa pengawasan dari data yang tidak terstruktur atau tidak berlabel. Juga dikenal sebagai <strong>deep neural learning</strong> atau <strong>deep neural network</strong>.</p><p>Bagaimana kita bisa mendapatkan mesin untuk belajar seperti manusia? berikut ini adalah contoh tentang bagaimana <em>Deep Learning</em> digunakan.</p><ol><li>Computer Vision</li><li>Language Recognition</li><li>Sentiment Analysis</li><li>Text Analysis</li><li>Image Recognition</li></ol><p>Setelah mengetahui tentang Deep Learning dan beberapa contoh kegunaannya. Saatnya mencoba implementasi Deep Learning dengan menggunakan <strong>Deep Learning Framework.</strong></p><blockquote>Pengertian Deep Learning Framework</blockquote><p>“<strong>Deep learning framework</strong> is an interface, library or a tool which allows us to build deep learning models more easily and quickly, without getting into the details of underlying algorithms. They provide a clear and concise way for defining models using a collection of pre-built and optimized components.”</p><p>Jadi Deep learning framework adalah <em>tools atau library </em>yang memudahkan dalam implementasi model deep learning secara lebih cepat dan mudah tanpa harus membuat dari awal baris kode pemrograman. Beberapa contoh dari Deep learning framework antara lain:</p><ol><li>TensorFlow</li><li>PyTorch</li><li>Caffe dan Caffe2</li><li>Keras</li><li>MXNet</li><li>CNTK</li></ol><p>Untuk bagian pertama ini akan fokus pada implementasi <em>Deep Learning </em>menggunakan TensorFlow dengan bahasa pemrograman Python.</p><blockquote>Apa itu TensorFlow?</blockquote><p>TensorFlow adalah salah satu Framework yang digunakan dalam <em>Deep Learning </em>yang merupakan ciptaan Google dan merupakan framework <em>d</em>eep <em>learning</em> yang paling terkenal. Framework ini merupakan <em>open source </em>sehingga kita dapat menggunakannya secara gratis.</p><p>Nama Tensorflow secara langsung berasal dari kerangka intinya: Tensor. Dalam Tensorflow, semua perhitungan melibatkan tensor. Tensor adalah vektor atau matriks n-dimensi yang mewakili semua jenis data. Semua nilai dalam tensor memiliki tipe data identik dengan bentuk yang diketahui (atau sebagian diketahui). Bentuk data adalah dimensi dari matriks atau array.</p><p>Pada TensorFlow, semua operasi dilakukan di dalam <strong>graph</strong>. Sedangkan <strong>Graph</strong> sendiri adalah seperangkat perhitungan yang terjadi secara berturut-turut. Setiap operasi disebut <strong>operation (op) node</strong> yang terhubung satu sama lain. Dalam <em>Machine Learning</em>, model dimasukkan dengan daftar objek yang disebut vektor fitur.</p><p>ektor fitur dapat berupa tipe data apa pun. Vektor fitur biasanya akan menjadi input utama untuk mengisi tensor. Nilai-nilai ini akan mengalir ke op node melalui tensor dan hasil dari operasi / perhitungan ini akan menciptakan tensor baru yang pada gilirannya akan digunakan dalam operasi baru. Semua operasi ini dapat dilihat dalam grafik.</p><p>Untuk lebih jelasnya mengenai TensorFlow, silakan lihat tautan berikut.</p><p><a href="https://www.tensorflow.org/">TensorFlow</a></p><p><strong>Representasi Tensor dengan Matrix</strong></p><p>Dalam TensorFlow, tensor adalah kumpulan vektor fitur (Array) dari n-dimensi. Contohhya adalah, jika kita memiliki matriks 2x3 dengan nilai dari 1 sampai 6:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/110/1*AJSkDBA1x9f94diBUlLUnw.png" /></figure><p>TensorFlow menggambarkan matriks ini sebagai berikut:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/127/1*eDdribk6L-L5gq7lr-nomg.png" /></figure><p><em>*Tensor dapat direpresentasikan dengan skalar atau dapat memiliki bentuk lebih dari tiga dimensi. Lebih rumit untuk memvisualisasikan level dimensi yang lebih tinggi.</em></p><h3><strong>Jenis-jenis Tensor</strong></h3><p>Dalam TensorFlow, semua perhitungan melewati satu atau lebih tensor. Tensor adalah objek dengan tiga properti:</p><ul><li><em>A unique label (name)</em></li><li><em>A dimension (shape)</em></li><li><em>A data type (dtype)</em></li></ul><p>Setiap operasi yang akan dilakukan dengan TensorFlow melibatkan manipulasi tensor. Ada empat tensor utama:</p><ol><li>tf.Variabel</li><li>tf. konstan</li><li>tf.placeholder</li><li>tf.SparseTensor</li></ol><p>TensorFlow works around 3 main components:</p><ul><li>Graph</li></ul><p>Semua operasi matematika (ops) dilakukan di dalam graph<em>, dan graph</em> sebagai proyek di mana setiap operasi dilakukan. <em>Node</em> mewakili ops ini, mereka dapat menyerap atau membuat tensor baru.</p><ul><li>Tensor</li></ul><p>Merupakan data (atau nilai) yang akan mengalir (flow) dalam <em>graph </em>dan merupakan <em>edge</em> dalam <em>graph</em>.</p><ul><li>Session</li></ul><p>Session akan menjalankan operasi dari graph.</p><p><strong>Implementasi TensorFlow dengan Python</strong></p><p>TensorFlow bekerja dengan pertama-tama mendefinisikan dan menggambarkan model secara abstrak, dan kemudian mewujudkannya dalam sebuah <em>session</em>. Deskripsi model adalah <em>“Computation Graph”. </em>Mari kita coba mengimplementasikan dengan contoh sederhana. Langkah pertama, mari kita buat sebuah <em>Tensor</em>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/370/1*T9-Qc_0mOYbKnpL294aHPw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/668/1*8ttraMC8GJDxejtIWhP2iw.png" /><figcaption><strong>Output</strong></figcaption></figure><h3>Operator Matematika TensorFlow</h3><p>TensorFlow berisi semua operasi dasar. Anda bisa mulai dengan yang sederhana. Anda akan menggunakan metode TensorFlow untuk menghitung kuadrat angka. Operasi ini mudah karena hanya satu argumen yang diperlukan untuk membangun tensor.</p><p>Berikut ini adalah daftar operasi yang umum digunakan. Idenya sama. Setiap operasi membutuhkan satu atau lebih argumen.</p><p><strong>tf.add (a, b)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/313/1*c9-Cv0A54Xu-_TcDkA02aA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/673/1*o98YjJ0VBKXXBgSe7e7RYA.png" /></figure><p><strong>tf.multiply (a, b)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/347/1*y1MJNz6TOl-dxtCOb9f3jA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/677/1*lJR_FGukTn-tCUSvjg4_FA.png" /></figure><p><strong>tf.sqrt (a)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/357/1*3uKlqHlcGf4sgKcQ-6INiw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/677/1*oWygGc_dPDioBQ-PWyYzOQ.png" /></figure><p><strong>tf.subtract (a, b)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/344/1*Rj1kYj6VvftnIHQQO4wnaw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/682/1*0tLvuBONtZE7-OqeO2ZlLg.png" /></figure><p><strong>tf.div (a, b)</strong></p><p><strong>tf.pow (a, b)</strong></p><p><strong>tf.exp (a)</strong></p><p><strong>Kesimpulan:</strong></p><p>TensorFlow adalah sebuah <em>tools, library</em> dari <em>artificial intelligence</em> yang merupakan open source. Ini memungkinkan pengembang untuk membuat <em>neural network</em> berskala besar dengan banyak lapisan. TensorFlow terutama digunakan untuk: Klasifikasi, Persepsi, dan Prediksi.</p><p>Tiga hal utama yang harus dipahami ketika mengimplementasikan TensorFlow adalah:</p><ol><li><em>Graph</em></li><li><em>Tensors</em></li><li><em>Session</em></li></ol><p>Tulisan bagian 2 akan membahas mengenai implementasi TensorFlow untuk <em>classification</em> dan <em>Image Recognition</em>.</p><p><strong>References:</strong></p><p>Deep Learning &amp; Artificial Neural Networks. Retrieved from <a href="https://machinelearningmastery.com/what-is-deep-learning/">https://machinelearningmastery.com/what-is-deep-learning/</a></p><p>Devopedia. 2019. “Deep Learning Frameworks.” Version 9, January 22. Accessed 2019–09–06. <a href="https://devopedia.org/deep-learning-frameworks">https://devopedia.org/deep-learning-frameworks</a></p><p>Suk, H.-I. (2017). An Introduction to Neural Networks and Deep Learning. Deep Learning for Medical Image Analysis, 3–24. doi: 10.1016/b978–0–12–810408–8.00002-x</p><p>TensorFlow Basics: Tensor, Shape, Type, Graph, Sessions &amp; Operators. Retrieved from <a href="https://www.guru99.com/tensor-tensorflow.html">https://www.guru99.com/tensor-tensorflow.html</a></p><p>What Is Deep Learning?: How It Works, Techniques &amp; Applications. Retrieved from <a href="https://www.mathworks.com/discovery/deep-learning.html">https://www.mathworks.com/discovery/deep-learning.html</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8b84c751ae34" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fake News Throughout History]]></title>
            <link>https://syahrier-wakid.medium.com/fake-news-throughout-history-ad2354ce4c0a?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/ad2354ce4c0a</guid>
            <category><![CDATA[hoax]]></category>
            <category><![CDATA[fake-news]]></category>
            <category><![CDATA[medium]]></category>
            <category><![CDATA[history]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Wed, 21 Aug 2019 12:50:47 GMT</pubDate>
            <atom:updated>2020-06-26T04:41:16.547Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4Nt5gc8Qm9yaR_lt3WHKkw.jpeg" /><figcaption>Image soruce: pentictonwesternnews.com</figcaption></figure><h3><strong>The Long History of Fake News</strong></h3><p>The term of “Fake News” become viral in the time of U.S Presidential Election on 2016. It becomes famous word after Donald J. Trump said it in the interview with some media, not to mention that he repeats it that words in his hundreds of his tweet, we can see it from his Twitter account @realDonaldTrump. He may popularize the term of Fake News, however long before Trump, who is now the President of the United States of America talk about fake news, the concept of fake news has been around for a long time (Posetti &amp; Matthews, 2018), it even goes way back in the ancient time before the newspaper era and of course before the digital era, to the story of Cleopatra and Antony (Sirianni, 1984) and (MacDonald, 2018).</p><p><strong>The Story of Mark Antony and Cleopatra</strong></p><p>Circa 44 BC, it starts of as the rivalry between Octavian, the heir of Julius Caesar, and his right-hand man, Marcus Antonius (Mark Antony). Both of them has a ferocious propaganda two win the war. His full name is Gaius Octavius. He is more famous with the nickname Octavian. he changed his name to Gaius Julius Caesar Octavianus, aka Octavian, upon being adopted by his great-uncle, Julius Caesar. In the biography of Augustus Caesar (Greenspan, 2014), it is known that Caesar’s death caused a fierce and long struggle between political and military leaders to occupy the seat of power. Octavian was involved in this struggle. They consider Octavian lightly not rivals that need to be taken into account. Whereas Octavian, feeling like a child of Caesar, wanted to take advantage of this situation.</p><p>With the careful plan, he tried to seize political victory by seeking the support of Caesar’s forces and appointed Mark Anthony as his main support because Mark Antony was Caesar’s closest friend. Mark Antony<em> </em>or Marcus Antonius Roman general under Julius Caesar. He is considered as the right-hand of Julius Caesar (History.com Editors, 2009). A series of battles in the following years could eliminate political opponents in order to seize power. In 36 BC Rome and many other areas were defeated divided by Mark Antony and Octavian. Mark Antony controls the eastern state and western Octavian. But the closeness and cooperation between the two of them did not last long. And the reason for their split is Cleopatra, queen of Egypt. the hostility broke out between Octavian and Antony, it is all because of Cleopatra. Roman historian Cassius Dio described Cleopatra as a woman of beauty. While the Greek historian Plutarch relates that Cleopatra beauty, was completely unmatched by any woman. In fact, Cleopatra’s beauty can fascinate everyone (Heritage Daily — Archaeology News, 2019)</p><p>Regarding the civil war, Octavian has the upper hand since he lives in the west (Italy), which can directly influence the senate in Rome. So, he used the piece of document that Octavian get a hold, which he called the will and testament of Antony. Even until now, it is debatable between scholars whether the document is real or not. But the point is, that piece of document was the main cause of the downfall of Antony. The document the contents of the document were Octavian’s propaganda to influence the senate in Rome that Antony prefers Cleopatra rather than the Roman empire. This way, the senate declared war to Antony, and in the end Antony and Cleopatra included were lost in the war which ironically led by Octavian.</p><p><strong>The Hoaxes (Fake News) History in The Early Modern Time</strong></p><p>It turns out hoax themselves have circulated since Johannes Gutenberg invented the printing press in 1439 ago (Posetti &amp; Matthews, 2018). Sadly, hoax in those days is even more dangerous than in the digital and internet era like today. As a result, circulating hoax are difficult to verify. Meanwhile, according to researcher Lynda Walsh in a book entitled Sins Against Science, hoax itself is an English term that began to enter since the industrial era or around 1808 (Castagnaro, 2009). However, as long as the word hoax itself is believed to have existed and appeared hundreds of years before century, as proved by the story of Mark Antony and Cleopatra.</p><p>On the other hand, in a book called the Museum of Hoaxes, Alexander Boese explained that the first hoax published was a fake almanac made by Isaac Bickerstaff aka Jonathan Swift in 1709 ago. At that time, he predicted the death of astrologer John Partridge. Then, in order to convince the public, Swift even made a false obituary about Partridge on the day predicted as the day of his death. In the process, he fabricated the information to embarrass Partridge in the public eye. As a result, Partridge stopped making astrological almanacs for up to six years after hoaxes circulated (The Museum of Hoaxes). Literary deception parodies phrases to become a myth that many people believe.</p><p>Hoax is considered a curse within the limits of achieving the myth of doctrine and propaganda over belief, all lies continue to evolve and infect the facts that occur. Here are some examples of literary hoaxes that have been recorded in history.</p><p><strong>The Great Moon Hoax</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/961/1*ItsB4QqwFeu1Lj_51hPaoQ.png" /><figcaption>The Great Moon Hoax by the tabloid <em>The</em> <em>Sun</em> from 1835. Depiction of moon surface. Source: <a href="https://en.wikipedia.org/wiki/Great_Moon_Hoax">Wikipedia</a></figcaption></figure><p>Another example is the Great Moon Hoax that begin at 1835, published in the New York Sun (Castagnaro, 2009) and (The Great Moon Hoax). On that year, the New York Sun newspaper launched the first of six articles announcing life’s findings on the moon with the title GREAT ASTRONOMICAL DISCOVERIES LATELY MADE BY SIR JOHN HERSCHEL, L.L.D. F.R.S. &amp;c. At the Cape of Good Hope. The article, then called The Great Moon Hoax, is allegedly a reissue of a journal called Edinburgh Journal of Science. In the article there is the name of Dr. Andrew Grant who was described as a colleague of Sir John Herschel, a famous astronomer at that time. In fact, Herschel had indeed been to Cape Town, South Africa, in January 1834 to establish an observatory with a new supermarket. However, according to Grant, Herschel has found new evidence of life forms on the moon, such as the form of fantastic animals such as unicorns and furry animals resembling humans with bat wings.</p><p>The article also details the geography of the moon, complete with large craters, large violet gem crystals, rivers, and dense vegetation. The New York Sun was founded in 1833, appeared in the public as a daily that had a low price and a style of narrative journalism that had a large portion. Since month hoax articles were published, their sales soared. The problem is that no single article is really a fact. In fact, the Edinburgh Journal of Science has stopped publishing it for years before being published. The article was mostly written by Richard Adams Locke, a reporter for Sun who had studied at Cambridge University. It was intended as a satire designed to make fun of serious speculation about life in space.</p><p>Something that failed to be captured by readers’ thinking at that time which can understandable because it was very hard to verify the authenticity of the news, well, no internet available and certainly no smartphones either. On September 16, 1835, New York Sun admitted that the article was a hoax. However, people can accept it and sales don’t go down. Surprisingly also, the daily publicity of the New York Sun soared to 19.360 per day. It even spread all over the US and Europe and became the hot topics in all parts of society. In term of journalism and news outlets, the story of The Great Moon Hoax is the first fake news that widely published (The Great Moon Hoax).</p><p><strong>The Balloon Hoax</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/183/1*b62vyUdvNNL5RIotMByKEQ.jpeg" /><figcaption>Image source <a href="https://onemorelibrary.com/">https://onemorelibrary.com</a></figcaption></figure><p>In 1835, the American poet and poet Edgar Allan Poe, published a famous hoax story; The <em>Unparalled Adventure of One Hans Pfaall</em> about the man who went to the moon using a hot air balloon and lived there for 5 years (Blacklock, 2015). But the echo is still inferior to the story of the moon hoax. In 1844, Poe wrote another article for The New York Sun which was the forerunner of the recursive nature and development of the definition of hoax we know today. In his article Poe told about exploration around Europe across the Atlantic Ocean in just three days using a hot air balloon by Monck Mason.</p><p>The newspaper article that tells the journey of a hot air balloon across the Atlantic was written by Poe as a true story. However, Poe finally admitted that the story was just a hoax. It is possible that Poe’s story was inspired by the hoax of the Great Moon Hoax published in the same newspaper nine years earlier (1835), claiming that astronomer John Herschel had seen humanoids winged on the moon using sophisticated new telescopes. (The Great Balloon-Hoax, 1850)</p><p>The story known as The Ballon-Hoax was first published by The Sun New York newspaper. Poe felt that the story of the Great Moon Hoax had plagiarized the basic idea of the story “The Unparalleled Adventure of One Hans Pfaall” which was a hoax written by Poe in The Sun newspaper, but the deception did not succeed in becoming famous until The Sun did not provide compensation at all for his Balloon story.</p><p><strong>The Sokal’s Affair</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/1*-u3TqHVadboqStUMrvKPkw.jpeg" /></figure><p>One of the famous names that can be referred to discuss the other side of the fake news or hoax is Alan Sokal, a professor of physics at New York University and professor of mathematics at University College London (Goldcare, 2013). His name is even enshrined in a term that is quite famous: Sokal’s hoax. In the 46th / 47th edition of Spring 1996, Social Text, a well-known journal in the United States, published a paper entitled Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity.</p><p>The 39-page paper was written by Alan Sokal, a professor of physics at New York University and professor of mathematics at University College London. In terms of titles, the paper looks ambitious. Two words that become important keywords in two scientific traditions, namely “hermeneutics” in cultural studies and “quantum gravity” in science, combine in a series of sentences beginning with a word: “Transgressing”. The contents of the paper are no less exciting. Sokal cited many post-modern thinkers cum structuralists such as Jacques Derrida, Jean-Francois Lyotard, Luce Irigaray, Gilles Deleuze, Felix Guattari, and Jacques Lacan. Sokal aggressively combined the thinker’s quotes with the latest findings of physicists such as Heisenberg, Einstein and Niels Bohr (Lezard, 2010).</p><p>Unexpectedly, a few weeks later, precisely on April 15, 1996, in an essay entitled Physicist Experiments with Cultural Studies published in the journal Lingua Franca, Alan Sokal revealed that his paper which aired on Social Text was merely a parody to ridicule postmodern thinkers. In other words, the paper is a hoax. The fact is that the paper was a hoax, he wrote it to test the standards of US humanities academic intellectuals. The formulation of the problem posed is quite simple. Sokal put it this way: “Will a well-known cultural study journal in America publish an article full of nonsense if (1) it seems good and (2) in line with the ideological pre-conception of journal editors?”. Sokal’s question was answered. The hoax paper was received and published by Social Text — founded by three academics, namely John Brenkman, Stanley Aronowitz, and Fredric Jameson (a widely quoted postmodern thinker in cultural studies).</p><p>According to Sokal, this happened because, the Social Text editor liked his article only because he had conclusions that were in accordance with the ideology of the editors, namely the content and methodology of postmodern science provided strong intellectual support for progressive political projects. What Sokal has done to this day is often referred to as evidence of how the academic community is not impermeable to information that is not tested. If a journal lead by experts is easily mislead, let alone ordinary people who are not accustomed to reviewing, testing or verifying information.</p><p>This experiment is a lesson that a well-crafted false information that mislead people can be created. This is very dangerous because the impact can be very broad and does not stop only in certain circles such as the academic community in the Sokal’s affair especially with the power of social media. This also raise a question about the structure of the paper that can deceive people. Write something the reader wish to believe, no matter if it is true or not, as long as your reader believe it, then you successfully creating hoax.</p><p>Then, entering the all-sophisticated twentieth century, begin the internet Era, hoax news often appeared through broadcast channels rather than print media. The reason is of course because it keeps up with the times, so the news delivery must be as fast as possible. As a result, many mass media did not clarify information before publishing it. One of the most famous hoaxes of the 20th century was when ABC television broadcasts and USA Today claimed that Russia planned to sell Vladimir Lenin’s body to boost state revenues (Museum Hoax) and (Trueheart, The Washington Post, 1991).</p><p>In addition, there is also one hoax that often circulates, namely about the threat of asteroids that will hit the earth to cause doomsday. In this case, NASA, in 2015, denied rumors of falling asteroids and caused massive damage to the earth. NASA explained that potentially dangerous asteroids had 0.01 percent impact on the earth over the next 100 years. “If there are large objects that will damage in September, of course we have acted now,” said NASA’s Near-Earth Object Manager Paul Chodas, August 2015. Of course, the hoax in this digital era is claimed to be more dangerous and frightening because it can be made very easily and the spread can be faster through the internet network. In fact, fake news have now spread to various segments of political, scientific, economic, social and entertainment life.</p><p><strong>References</strong></p><p>Anonymous. History.com Editors. (2009, December 16). Mark Antony. Retrieved April 2, 2019, from <a href="https://www.history.com/topics/ancient-history/mark-antony">https://www.history.com/topics/ancient-history/mark-antony</a></p><p>Anonymous. The Great Moon Hoax. (n.d.). Retrieved April 2, 2019, from <a href="http://hoaxes.org/archive/permalink/the_great_moon_hoax">http://hoaxes.org/archive/permalink/the_great_moon_hoax</a></p><p>Blacklock, M. (2015, August 05). Top 10 literary hoaxes. Retrieved April 3, 2019, from <a href="https://www.theguardian.com/books/2015/aug/05/top-10-literary-hoaxes-mark-blacklock">https://www.theguardian.com/books/2015/aug/05/top-10-literary-hoaxes-mark-blacklock</a></p><p>Castagnaro, M. (2009), Embellishment, Fabrication, and Scandal: Hoaxing and the American Press. Carnegie Mellon University. Ph.D. Dissertation. [pdf] Retrieve April 2, 2019 from <a href="https://pqdtopen.proquest.com/doc/304849843.html?FMT=ABS">https://pqdtopen.proquest.com/doc/304849843.html?FMT=ABS</a></p><p>Greenspan, J. (2014, August 19). 8 Things You May Not Know About Augustus. Retrieved April 2, 2019, from <a href="https://www.history.com/news/8-things-you-may-not-know-about-augustus">https://www.history.com/news/8-things-you-may-not-know-about-augustus</a></p><p>Goldcare, B. (2003, June 05). The Sokal affair. Retrieved March 7, 2019, from <a href="https://www.theguardian.com/science/2003/jun/05/badscience.research">https://www.theguardian.com/science/2003/jun/05/badscience.research</a></p><p>Heritage Daily — Archaeology News. (2019). Was Cleopatra beautiful? The archaeological evidence. Retrieved April 2, 2019, from <a href="https://www.heritagedaily.com/2015/08/was-cleopatra-beautiful-the-archaeological-evidence/92015">https://www.heritagedaily.com/2015/08/was-cleopatra-beautiful-the-archaeological-evidence/92015</a></p><p>MacDonald, E. (2018, September 18). The fake news that sealed the fate of Antony and Cleopatra. Retrieved from <a href="https://theconversation.com/the-fake-news-that-sealed-the-fate-of-antony-and-cleopatra-71287">https://theconversation.com/the-fake-news-that-sealed-the-fate-of-antony-and-cleopatra-71287</a></p><p>Posetti, J., &amp; Matthews., A. (2018). A Short Guide to History of Fake News and Disinformation. A Learning Module for Journalists and Journalism Educators. ICFJ (International Center for Journalist). [pdf] Retrieve April 02, 2019 from <a href="https://www.icfj.org/">https://www.icfj.org</a></p><p>Sirianni, F. (1984). WAS ANTONY’S WILL PARTIALLY FORGED? L’Antiquité Classique, 53, 236–241. Retrieved from <a href="http://www.jstor.org/stable/41657427">http://www.jstor.org/stable/41657427</a></p><p>Trueheart, C. (1991, November 07). IS LENIN FOR SALE? OF CORPSE NOT! Retrieved April 3, 2019, from <a href="https://www.washingtonpost.com/archive/lifestyle/1991/11/07/is-lenin-for-sale-of-corpse-not/0ced579a-0df8-43ca-a76d-ead58be28a6d/?utm_term=.a8709ba34057">https://www.washingtonpost.com/archive/lifestyle/1991/11/07/is-lenin-for-sale-of-corpse-not/0ced579a-0df8-43ca-a76d-ead58be28a6d/?utm_term=.a8709ba34057</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ad2354ce4c0a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why are humans so difficult to detect Fake News?]]></title>
            <link>https://syahrier-wakid.medium.com/why-are-humans-so-difficult-to-detect-fake-news-d8d87a4cdafa?source=rss-b4fd4860eafd------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8d87a4cdafa</guid>
            <category><![CDATA[fake-news]]></category>
            <category><![CDATA[psychology]]></category>
            <category><![CDATA[deceived]]></category>
            <category><![CDATA[lies]]></category>
            <category><![CDATA[fake-news-stories]]></category>
            <dc:creator><![CDATA[Syahrier Wakid]]></dc:creator>
            <pubDate>Tue, 06 Aug 2019 10:24:32 GMT</pubDate>
            <atom:updated>2020-06-25T05:41:19.675Z</atom:updated>
            <content:encoded><![CDATA[<h3>Why Humans Are So Difficult to Detect Fake News?</h3><p>The recent studies (Frank and Feeley, 2003) and (Frank et al, 2004) shows that Human actually bad at judging whether some statements, articles or news is fake or not, in brief human are bad at detecting lie. In the first research which is the participant is a student from Rutgers University, it shows that only 54% accuracy rates, it shows that human slightly higher than the probability to guess between lie or truth. In the second research with participant from law enforcement personnel, witness credibility assessment experts show only slightly higher than the first research experiment which is from student participation.</p><p>Study conducted by (Qiu et al, 2017) found that one of the factors that made fake news widely spread was because of the limitations of the human brain in receiving information. The study shows that when someone receives an overloaded new information, they tend to rely on a handling mechanism that is not appropriate to distinguish a good or bad news, and finally choose a news that is popular and does not pay attention to the quality aspects of the news. Another research (Keersmaecker and Roets, 2017) find out how someone adjusts their attitude after knowing the important information they receive is incorrect information, and also to determine the extent to which cognitive abilities influence the evaluation of information obtained. The subject of this study was presented on a negative information about someone, after that was given an explanation that the negative information turned out to be wrong.<br> <br>The results of the research show that a person with low cognitive abilities, less responsive (lack of response) to correct information correction, exposure to old information that is wrong has a strong influence on them. Specifically, when individuals with low cognitive abilities know that their attitudes toward the target audience (in this case the person being reported) are based in part on incorrect negative information, they adjust their evaluations to the target, but the numbers are smaller than individuals with high cognitive abilities. Individuals with low cognitive abilities who evaluate the target are still more negative than evaluating individuals who do not receive any information. Individuals with high cognitive abilities make more appropriate attitudes after knowing that the negative information provided is wrong.</p><blockquote>Why us as a human cannot separating between fake news and true news?</blockquote><p>Because the human brain processes information in a biased way. There are many theories that suggest human bias tendencies (Hutson, 2017). One of them is what evolutionary psychologists call <strong>“positive bias”</strong> (Forgas, 2018), where the brain is designed to assume that most people act in an honest and sincere way. Positive bias, besides allowing us to be preferred by others, also makes the threshold of acceptance lower or not critical, so it is easier to believe in a lie. Of course, positive bias does not apply continuously because we only apply it to people who are known or worthy of us. There are also times when a person is aware, even if only intuitively, that other people are manipulating it for personal purposes.</p><p>Because, every human being has a fear of ignorance. In addition, a Nobel winner, psychologist Daniel Kahneman, in his lecture article (Kahneman, 2013), he said that a human had evolved with two concepts of thought, namely the first intuitive system of thought like our ancestors, and the second, analytic and more rational system of thought. Another study from Paul Ekman, an emeritus professor of psychology at the University of California, San Francisco, even though human is able to think rationally, if the source is a known person, someone can still be easily deceived because of a desire to believe to the person we know personally or close to us. Even to give a second chance (Ekman, 2017). This desire factor refers to a psychological construction called <strong>“cognitive dissonance”</strong>, that is when you convince yourself to believe even if it is not true on things you want to believe, and ignore facts that you do not want to believe.</p><p>We tend to trust people we know. This means that the more we see people close to us talking about an opinion, the more we believe the truth of that opinion. Even though the people who close to us are not experts, that does not affect our level of trust. When the statement is then relied repeatedly, for example, through a television program, share it in social media, this will lead to the illusion that the opinion is popular and more acceptable than the fact. And in the end, we will believe it as a truth.</p><p>There are other theories that explain why people are easily lied to. First, confirmation bias (Hesma, 2015). That is, the action seeks justification to confirm the belief, by ignoring (valid) information that is contradictory and instead chooses questionable information. Confirmation bias was found in one study as a trigger for hoax news. Second, <strong>“cognitive misers”</strong> or cognitive misers (Robson, 2016), which describe our brains prefer to trust intuition rather than analysis as an instant way without hassle. A well-known example of this theory is a condition called <strong>“the illusion of Moses”</strong> summed up based on survey questions:</p><blockquote>“How many types of animals did Moses bring to his ark?”</blockquote><p>Illusion Moses explained that we are so easy to ignore the details of a statement, because there are other things that are considered more interesting. About 10–50 percent of the people surveyed failed to realize that the one who owned the ark was Noah, not Moses. So, even though we know we have to trace facts and evidence, we still use feelings based on self-assessment of what is right and wrong.</p><p>This deceptive effect of the illusion of Moses, even though at first it might have raised doubts, would be believed to be absolute truth when repeated. Besides the explanation above, there are many factors that make us easier to deceive. Mood for example, can affect the way we process lies (Forgas and East, 2008). People who are happy are even more reckless than people who are angry when lied to. Negative moods help us be more focused and careful in choosing information. Then, people who feel smarter than others also tend to be more easily lied to. The reason, because they feel they have more knowledge, so they tend to be reluctant to pay attention to the facts.</p><p>By getting the fact that humans are bad detectors for fake news. Fake news creators utilize the goodness of humans who have the nature of empathy and rely on emotions. In the end, we cannot rely on ourselves to detect fake news as we can easily be deceived by lies.</p><p><strong>References</strong></p><p>Ekman, D. K. (2017, December 21). Opinion | The Science of ‘Inside Out’.</p><p>Frank, M. G., &amp; Feeley, T. H. (2003). To Catch a Liar: Challenges for Research in Lie Detection Training.</p><p>Frank, M. G., Feeley, T. H., Paolantonio, N., &amp; Servoss, T. J. (2004). Individual and Small Group Accuracy in Judging Truthful and Deceptive Communication.</p><p>Forgas, J. P. (2018, December 10). Why are some people more gullible than others?.</p><p>Forgas, J. P., &amp; East, R. (2008). On being happy and gullible: Mood effects on skepticism and the detection of deception.</p><p>Hutson, M. (2017, April 04). Living a Lie: We Deceive Ourselves to Better Deceive Others.</p><p>Kahneman, D. (2003). A perspective on judgment and choice: Mapping bounded rationality.</p><p>Keersmaecker, J. D., &amp; Roets, A. (2017). ‘Fake news’: Incorrect, but hard to correct. The role of cognitive ability on the impact of false information on social impressions.</p><p>Qiu, X., Oliveira, D. F., Shirazi, A. S., Flammini, A., &amp; Menczer, F. (2017). Limited individual attention and online virality of low-quality information.</p><p>Robson, D. (2016, March 24). Future — Why are people so incredibly gullible?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d8d87a4cdafa" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>