<?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 Reid Francis on Medium]]></title>
        <description><![CDATA[Stories by Reid Francis on Medium]]></description>
        <link>https://medium.com/@reidfrancis?source=rss-896c87d7252a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*7Ho7Jb2kvxY7aLTniJyOzg.jpeg</url>
            <title>Stories by Reid Francis on Medium</title>
            <link>https://medium.com/@reidfrancis?source=rss-896c87d7252a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 19:06:45 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@reidfrancis/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[Automatically translate your documents with Watson Language Translator]]></title>
            <link>https://medium.com/@reidfrancis/automatically-translate-your-documents-with-watson-language-translator-ffcaab91f629?source=rss-896c87d7252a------2</link>
            <guid isPermaLink="false">https://medium.com/p/ffcaab91f629</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[box]]></category>
            <category><![CDATA[translation]]></category>
            <category><![CDATA[watson]]></category>
            <category><![CDATA[machine-translation]]></category>
            <dc:creator><![CDATA[Reid Francis]]></dc:creator>
            <pubDate>Thu, 14 Feb 2019 16:30:08 GMT</pubDate>
            <atom:updated>2019-02-14T16:30:08.129Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*_FqX50oJ5ABQEnKQ" /><figcaption>Photo by <a href="https://unsplash.com/@rawpixel?utm_source=medium&amp;utm_medium=referral">rawpixel</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>As your business grows, shouldn’t you have a way to grow and scale your language capabilities? Human translation alone can be expensive and time-consuming. That is why organizations are beginning to adopt machine translation capabilities. However, humans are generating text data across a wide variety of file formats that these machine translation capabilities do not support outright. This will require developers to build in functions for generating new documents for business professionals, and slowing down the process (or hurting the business case) for adopting machine translation capabilities.</p><p>Document Translation (Beta) with Watson Language Translator reduces your development time and complexity. Instead of focusing on how to seamlessly format documents once they are translated, let Watson do it. Now you can get translated documents into the hands of professionals more rapidly, improving the translation process and business efficiency.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0gcwiZxp07jNF3F2K2zCZg.png" /><figcaption>Using Document Translation with Watson Language Translator</figcaption></figure><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FDeT-iu8IXKM%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DDeT-iu8IXKM&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FDeT-iu8IXKM%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/0682fc2cbfd4c3c3683a150c008ab46d/href">https://medium.com/media/0682fc2cbfd4c3c3683a150c008ab46d/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k5Co0_1tGEMiYPzTZl0WEg.jpeg" /></figure><p>Companies like <a href="https://www.ibm.com/watson/box/">Box are already embedding document translation to automatically translate documents and preserve file formatting within their platform.</a></p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FupS61k7jcN0%3Fstart%3D112%26feature%3Doembed%26start%3D112&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DupS61k7jcN0&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FupS61k7jcN0%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="640" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4d28aaed375c2cab0df3b4ea61b13518/href">https://medium.com/media/4d28aaed375c2cab0df3b4ea61b13518/href</a></iframe><p><strong>Learn More:</strong></p><p><a href="https://www.ibm.com/watson/services/language-translator/">Product Page</a>| <a href="https://console.bluemix.net/docs/services/language-translator/translating-documents.html#document-translator-tutorial">Document Translation (Beta)</a> | <a href="https://console.bluemix.net/docs/services/language-translator/index.html#about">Documentation </a>| <a href="https://www.ibm.com/watson/services/language-translator/">Sample Apps</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ffcaab91f629" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Developers! Get ready for Black Friday with AI]]></title>
            <link>https://medium.com/@reidfrancis/developers-get-ready-for-black-friday-with-ai-57235994556?source=rss-896c87d7252a------2</link>
            <guid isPermaLink="false">https://medium.com/p/57235994556</guid>
            <category><![CDATA[search]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[getting-started]]></category>
            <category><![CDATA[classification]]></category>
            <category><![CDATA[tutorial]]></category>
            <dc:creator><![CDATA[Reid Francis]]></dc:creator>
            <pubDate>Tue, 20 Nov 2018 20:53:19 GMT</pubDate>
            <atom:updated>2018-11-27T20:02:08.876Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1t9ZuBAx7DAsQh40" /><figcaption>Photo by <a href="https://unsplash.com/@fancycrave?utm_source=medium&amp;utm_medium=referral">Fancycrave</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><em>[Shout out to </em><a href="https://www.linkedin.com/in/kyle-babinowich-92675528/"><em>Kyle Babinowich</em></a><em> for helping me write this post!]</em></p><p>Black Friday is just around the corner, and if you are like me, that means slipping into a food coma from Turkey Day leftovers and building a large online shopping cart of useless items on Amazon (I personally avoid the brick and mortar store madness). As exciting as this might sound, I always feel like I could be doing something more productive with my time. How about experimenting with AI?</p><p><a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> (NLC) is the easiest way to get started with text classification. Let’s take a look at three use cases to help e-commerce websites and retail applications make some significant upgrades with natural language processing (NLP) and machine learning (ML) capabilities. This post is going to cover the following topics and in each section there are sample data set links to help you get started trying out NLC:</p><ul><li>Improving search with intent classification for common queries</li><li>Product categorization for search and improved cross-sell</li><li>Categorizing product reviews to improve user experiences and gain customer insights</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*zCTHVKMcSOUFfFGe" /><figcaption>Photo by <a href="https://unsplash.com/@edhoradic?utm_source=medium&amp;utm_medium=referral">Edho Pratama</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Intent Classification for Common Search Queries</h3><p>Raise your hand if you get tired of continuously having to apply filters to a retail page because the search results are subpar. Why do websites put us through this tedious experience? If their data is categorized for filters, shouldn’t it be searchable? You would think so. Traditional search techniques are great at finding matches to keyword searches but not great at interpreting question intent or short queries. That is where NLP and ML can come in handy.</p><p>One of the most common use cases for NLC is intent classification. The service excels at deciphering intention in order to convert unstructured questions into structured queries. This is a core component that underlies chatbot technology (check out <a href="https://www.ibm.com/watson/ai-assistant/">Watson Assistant</a>). Intent classification is different than more traditional text classification because user queries are typically very short, ambiguous, and may belong to multiple categories (<a href="http://users.dsic.upv.es/~prosso/resources/HernandezEtAl_CERI12.pdf">check this paper out</a>). This requires services to leverage different parameters and techniques when tackling problems like document classification compared to intent classification. Additionally, using intent classification is not solely dependent on the frequency of keywords, or other elements, that are common in traditional search techniques, and are able to provide more precise, intuitive responses by using supervised learning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G5LsgKANd7AAkkm-Lq5eMw.png" /><figcaption>Use NLC to classify intent for search queries and populate are cache</figcaption></figure><p>In the above image you can see a scenario where NLC is used to populate the search cache for new search queries and the results of the classification are stored as part of the search query.</p><blockquote>“Caching is extremely important because you can get dramatic performance improvements without a lot of effort in certain situations.”</blockquote><blockquote>— <a href="https://medium.com/u/2bf02fddc3f8">Alex Landau</a>, <a href="https://hackernoon.com/why-do-we-cache-b24920e1e903">Why Do We Cache?</a></blockquote><h4><strong>Let me give you an example:</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*YitsM9c7BBScVBzn" /><figcaption>Photo by <a href="https://unsplash.com/@rawpixel?utm_source=medium&amp;utm_medium=referral">rawpixel</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><a href="https://www.ibm.com/case-studies/c733086k42590b22">1DOC3 S.A.S</a>, an online health portal company in South America, needed a more efficient way to process user queries as its online medical resource became increasingly popular. The escalating problem was the means of classifying all the knowledge that had been accumulated. With roughly 40,000 health queries coming in every month, 1DOC3’s leadership team calculated it would need 30 physicians just to classify the information. Needless to say, such an exercise would be a profound waste of their time and expertise, and of company resources.</p><p>The company created an interactive health advice platform using <a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> to help funnel inquiries, with increasing efficiency, toward the answers. Each time it encounters a query not answered before, the system directs it to the appropriate doctor, whose personal — and personalized — answer adds another layer to the knowledge base.</p><h4><strong>Here are some results…</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*VQXmJ6j9oU3CHJVgqjRG4A.png" /><figcaption>1DOC3 benefits from Watson NLC</figcaption></figure><p><strong>Helpful data set to get started with intent classification</strong></p><ul><li><a href="https://data.world/crowdflower/ecommerce-search-relevance">E-commerce search relevance</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Am3mc6Cf_fu4t80l" /><figcaption>Photo by <a href="https://unsplash.com/@mrleecanburn?utm_source=medium&amp;utm_medium=referral">mr lee</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3><strong>Product Categorization for Augmented Search &amp; Improved Cross-Sell</strong></h3><p>Imagine the last time you searched for a product using a retail company’s search bar. You likely searched something like “iPhone XS” or “Playstation 4”, and got a list of items that had those keywords in it, maybe ranked in some order, hopefully with the item that you really wanted to see at the top. What if in addition to that keyword search, we want a simple search of “iPhone XS” to give us not only that item first, but perhaps a whole digital store of Apple phone accessories like Airpods, charging cables, and cases? Or maybe we that search for “Playstation 4” to return a selection of consoles, controllers, or games that are currently on doorbuster sales for Black Friday.</p><p>With <a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> we can train a series of product categorization engines that add descriptive metadata such as brand name, type, or price and use that metadata to suggest other products that the consumer would likely be interested in. This allows the retail company to provide an intelligent match for the search query with more metadata to associate with a particular product index. It also opens up the possibility to give the user similar or complementary products that the company <em>wants</em> the customer to see in a controlled, tailored experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gITWDiFK255t4ztVSzt5Bg.png" /><figcaption><em>Use NLC for product categorization and improving cross sell</em></figcaption></figure><p>How would something like this differ from the traditional “Customers Who Bought This Item Also Bought” type of items that can be displayed next to a product? “Customers Who Also Bought” recommendations looks at historical purchases to display other results. Instead of lumping products by product type or brand, we merely display other results by things that people happen to purchase at the same time. This is classic collaborative filtering.</p><blockquote>“The most basic models for recommendations systems are collaborative filtering models which are based on assumption that people like things similar to other things they like, and things that are liked by other people with similar taste.”</blockquote><blockquote>— <a href="https://medium.com/u/2f052b054cee">Prince Grover</a>, <a href="https://towardsdatascience.com/various-implementations-of-collaborative-filtering-100385c6dfe0">Various Implementations of Collaborative Filtering</a></blockquote><p>While a useful concept for cross-sell, it doesn’t allow for the retail company to frame the purchasing experience the way they would like to. Some people may disagree with the concept of company curated recommendations, but think how successful <a href="https://www.creativedisplaysnow.com/six-inspired-examples-of-sales-boosting-grocery-store-end-caps/">grocery store end caps</a> are. With <a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> online retailers can easy duplicate the success of grocery store end caps in addition to the consumption bundling practice of “Customers Who Also Bought”.</p><p><strong>Try categorizing products for yourself!</strong></p><p>We glossed over the concept of using these classifiers to give a more robust index to more easily organize new products and product better results, but it is a huge point of value. If we churn every query and every product through these NLC categorization engines, we can use the additional metadata for improved search results. Instead of just matching on input keyword, to say, product name, we can match input keyword, input product category, input product cost, input product brand, etc. to our library of products.</p><p>An example of this is what was built in our <a href="https://github.com/erichensleyibm/NLC_product_classifier-demo">product categorization</a> sample application. In it, we demonstrate how <a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> allows an online marketplace to further optimize their inventory management and cataloging system by classifying products based on their description. <a href="https://github.com/erichensleyibm/NLC_product_classifier-demo">Fork the code</a> or <a href="https://erichensley-nlc-demo.mybluemix.net/">check out the demo</a> to try it yourself!</p><p><strong>Helpful data sets to get started with product categorization</strong></p><ul><li><a href="https://data.world/datafiniti/electronic-products-and-pricing-data">Electronic products data</a></li><li><a href="https://data.world/epa/cpcat-chemical-and-product-categories">Chemical and product categories</a></li><li><a href="https://www.kaggle.com/PromptCloudHQ/flipkart-products">Flipkart Products</a></li><li><a href="https://www.kaggle.com/PromptCloudHQ/all-jc-penny-products/data">JCPenney Products</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8SipzaHKRa3rdC12" /><figcaption>Photo by <a href="https://unsplash.com/@alvaroreyes?utm_source=medium&amp;utm_medium=referral">Alvaro Reyes</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Categorizing product reviews to improve user experiences and gain customer insights</h3><p>One of the more common text classification patterns I’ve seen is analyzing and labeling customer reviews. Understanding unstructured customer feedback enables organizations to make informed decisions that’ll improve customer experiences or identify issues faster. <a href="https://en.wikipedia.org/wiki/Sentiment_analysis">Sentiment analysis</a> is perhaps one of the most common text classification use cases, as it empowers businesses to understand the voice and tone of their customers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kzRTNuh1FswgxLBGhaFIIQ.gif" /></figure><p>Companies like Amazon use a variety of techniques to improve the customer experience for their users. The above gif demonstrates how Amazon provides categories for users to evaluate similar reviews. By displaying the most common categories, Amazon allows its customers to determine if the product he/she is reviewing is reviewed in a positive way. As you can see, although there is a category called ‘easy to use’ non of the top reviews under that category mention ‘easy to use’ as a single phrase. This is the power of using text classification. Amazon is able to define its categories from mining millions of reviews and then train its system to organize reviews into those categories.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*0dUhXHcIjcWuKpDG" /><figcaption>Photo by <a href="https://unsplash.com/@rawpixel?utm_source=medium&amp;utm_medium=referral">rawpixel</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Now let’s think about how companies can use this information internally. Companies can categorize reviews and count the occurrences of each category, or simply monitor the average sentiment across reviews over a period of time. This allows companies to monitor customer feedback at scale so that they can identify trends and issues as they appear within the reviews. Although this might be less important for a platform like Amazon, think about what a single retailer might be able to do with the customer feedback from their website. Rapid analysis of customer feedback allows companies to be more agile, innovate, and create overall better products.</p><p><strong>Want categorize reviews for yourself?</strong></p><p>Check out my recent tutorial: <a href="https://medium.com/ibm-watson/get-started-with-ai-in-15-minutes-28039853e6f3">Get Started with AI in 15 Minutes Using Text Classification on Airbnb Reviews</a>. In it, I show you how to build two classifiers to identify sentiment and categorize Airbnb reviews.</p><p><strong>Helpful datasets</strong></p><ul><li><a href="https://data.world/datafiniti/wine-beer-and-liquor-reviews">Wine, beer, and liquor reviews</a></li><li><a href="http://insideairbnb.com/get-the-data.html">Insideairbnb.com</a></li><li><a href="http://jmcauley.ucsd.edu/data/amazon/">Amazon review data</a></li></ul><h3>Helpful Links for Watson Natural Language Classifier</h3><p><a href="https://www.ibm.com/watson/services/natural-language-classifier/">Product Page</a> | <a href="https://console.bluemix.net/docs/services/natural-language-classifier/getting-started.html#natural-language-classifier">Documentation</a> | <a href="https://console.bluemix.net/docs/services/natural-language-classifier/sample-applications.html#sample-apps">Sample apps and code</a> | <a href="https://console.bluemix.net/apidocs/natural-language-classifier">API Reference</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=57235994556" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Get Started with AI in 15 minutes by Building Text Classifiers on Airbnb Reviews]]></title>
            <link>https://medium.com/@reidfrancis/get-started-with-ai-in-15-minutes-28039853e6f3?source=rss-896c87d7252a------2</link>
            <guid isPermaLink="false">https://medium.com/p/28039853e6f3</guid>
            <category><![CDATA[tutorial]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[classification]]></category>
            <category><![CDATA[development]]></category>
            <dc:creator><![CDATA[Reid Francis]]></dc:creator>
            <pubDate>Wed, 07 Nov 2018 17:01:02 GMT</pubDate>
            <atom:updated>2018-11-20T21:54:07.039Z</atom:updated>
            <content:encoded><![CDATA[<h3>Get Started with AI in 15 Minutes Using Text Classification on Airbnb reviews</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RrvNCnXaygybOKn9ovgoTQ.jpeg" /><figcaption>Build two text classifiers in 15 minutes</figcaption></figure><p><a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> (NLC) is a text classification (aka text categorization) service that enables developers to quickly train and integrate <a href="https://en.wikipedia.org/wiki/Natural_language_processing">natural language processing</a> (NLP) capabilities into their applications. Once you have the training data, you can set up a classification model (aka a classifier) in 15 minutes or less to label text with your custom labels. In this tutorial, I will show you how to create two classifiers using publicly available <a href="http://insideairbnb.com/get-the-data.html">Airbnb reviews</a> data.</p><p>One of the more common text classification patterns I’ve seen is analyzing and labeling customer reviews. Understanding unstructured customer feedback enables organizations to make informed decisions that’ll improve customer experience or resolve issues faster. <a href="https://en.wikipedia.org/wiki/Sentiment_analysis">Sentiment analysis</a> is perhaps one of the most common text classification cross-industry use cases, as it empowers businesses to understand voice and tone of their customers. However, companies also need to organize their data into categories that are specific to their business. This often requires data scientists to build custom machine learning models. With NLC, you can build a custom model in minutes without any machine learning experience.</p><h3>Training data</h3><p>To obtain training data, I went to <a href="http://insideairbnb.com/get-the-data.html">insideairbnb.com</a> and downloaded the <a href="http://data.insideairbnb.com/united-states/tx/austin/2018-10-18/data/reviews.csv.gz">‘reviews.csv.gz’</a> file from Austin, Texas. This file contains thousands of real reviews from Airbnbs in Austin.</p><p>Next, I defined my labels. I decided to build two classifiers one for categorizing the reviews and the other for sentiment. It was best to separate the training data for each and create separate classifiers in order to achieve the highest accuracy possible. The labels I defined are below:</p><ul><li><strong>Category Classifier:</strong> Environment, Location, Cleanliness, Hospitality, Noise, Amenities, Communication, Other</li><li><strong>Sentiment Classifier: </strong>Positive, Neutral, Negative</li></ul><p>Both sets of training data only contain 219 rows (examples). That isn’t a lot of examples in the grand scheme of things. However, one of the benefits of <a href="https://www.ibm.com/watson/services/natural-language-classifier/">Watson Natural Language Classifier</a> is that it works better on smaller sets of examples. Feel free to continue to add to the training data once you have downloaded the file to further improve the accuracy!</p><h3>Training the Classifiers</h3><p>In this tutorial, I will be using Watson Studio. If you would prefer to use the API directly, check out the <a href="https://console.bluemix.net/docs/services/natural-language-classifier/getting-started.html#natural-language-classifier">documentation</a>.</p><p>Create an instance of NLC and launch the tooling (Note: if you get lost, please refer to the embedded video at the bottom of this post):</p><ol><li>Go to the <a href="https://console.us-east.bluemix.net/catalog/services/natural-language-classifier">Natural Language Classifier</a> page in the IBM Cloud Catalog.</li><li>Sign up for a free IBM Cloud account or log in.</li><li>Click <strong>Create</strong>.</li><li>Once an instance is created, you will be taken to the below screen. Click <strong>Launch tool </strong>to open the tooling in Watson Studio.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UrKk9bR5euq7lJXiL-K4LA.png" /><figcaption>Open tooling from IBM Cloud Catalog</figcaption></figure><h3>Train your classifier</h3><ol><li>Download the training data. <em>Two columns is all you need! That’s how easy it is to train a classifier in NLC! </em><a href="https://ibm.box.com/s/jp4fkvk47tffj269y6cqne7whz5cyy99"><em>Download here!</em></a></li><li>Click “Create Model” to start building your classifier(s).</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0XDmLjXja3PGmyvgVn-79A.png" /><figcaption>Begin creating your classifier</figcaption></figure><ol><li>Next, you’ll need to create a project in Watson Studio. If you do not have an instance of Watson Studio created then you will need to provision a one on the Lite plan.</li><li>After you have provisioned your instance of Watson Studio, refresh the page and name your Watson Studio project. Then click “Create” in the bottom right hand corner.</li><li>Upload the training data for either the Categories or Sentiment</li><li>Click <strong>Train Model </strong>(Training will take approximately 5-10 minutes for each classifier)</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xbvevMH4S-8A5S6cK88SrQ.gif" /><figcaption>Uploading training data and training a classifier</figcaption></figure><h3>Testing your classifier</h3><ol><li>Now that training is done, you can test your classifier!</li><li>Click into your classifier and go to the <strong>Test </strong>page. Enter any text and see how Watson classifies it. The classifier works best when using actual Airbnb reviews — so test it out with data from <a href="http://insideairbnb.com/get-the-data.html">insideairbnb.com</a>. If the classifier makes a mistake, simply click <strong>Edit and Retrain</strong> in the top right corner and add more training examples to your training data. You’ll be classifying Airbnb reviews in no time!</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kcIDGy2Sy7i-dPKnNjZBHQ.gif" /></figure><p>Want to hook your classifiers up to a user interface? Check out the <a href="https://github.com/watson-developer-cloud/natural-language-classifier-nodejs">Github repo for the Natural Language Classifier demo</a>. This repo will give you the Node.JS for the <a href="https://natural-language-classifier-demo.ng.bluemix.net/">NLC demo</a> so you can hook your classifiers up to a simple user experience.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fb56utGwDvwg%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Db56utGwDvwg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fb56utGwDvwg%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/991af3e18cf4d59998e6815fc5befb6a/href">https://medium.com/media/991af3e18cf4d59998e6815fc5befb6a/href</a></iframe><h3>Helpful Links</h3><p><a href="https://www.ibm.com/watson/services/natural-language-classifier/">Product Page</a> | <a href="https://console.bluemix.net/docs/services/natural-language-classifier/getting-started.html#natural-language-classifier">Documentation</a> | <a href="https://console.bluemix.net/docs/services/natural-language-classifier/sample-applications.html#sample-apps">Sample apps and code</a> | <a href="https://console.bluemix.net/apidocs/natural-language-classifier">API Reference</a></p><p>Want to see what else Watson can do with Airbnb reviews? Check out the new demo for <a href="https://www.ibm.com/watson/services/discovery-3/demo/index.html#demo">Watson Discovery Service</a>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=28039853e6f3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>