<?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 Allen Kim on Medium]]></title>
        <description><![CDATA[Stories by Allen Kim on Medium]]></description>
        <link>https://medium.com/@akim88?source=rss-a4873189b53f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*jJngWB2lRr7gd9Bp.jpg</url>
            <title>Stories by Allen Kim on Medium</title>
            <link>https://medium.com/@akim88?source=rss-a4873189b53f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 16:05:19 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@akim88/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[What is Semantic Search?]]></title>
            <link>https://medium.com/graphiq-engineering/what-is-semantic-search-71808207cfcc?source=rss-a4873189b53f------2</link>
            <guid isPermaLink="false">https://medium.com/p/71808207cfcc</guid>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[semantics]]></category>
            <category><![CDATA[search]]></category>
            <dc:creator><![CDATA[Allen Kim]]></dc:creator>
            <pubDate>Tue, 17 May 2016 23:20:03 GMT</pubDate>
            <atom:updated>2016-05-20T01:17:15.164Z</atom:updated>
            <content:encoded><![CDATA[<h4><em>An Intro to the Changing Landscape of Search Technology</em></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KQEOXaH6svbprxEr11Htiw.png" /><figcaption>Graphiq Search</figcaption></figure><p>Let’s say you’re at an electronics store looking to upgrade your smartphone and a salesperson asks you, “Are you an Apple fan?” While she could be referencing the fruit, it’s much more likely she’s referring to the company that makes the <a href="https://smartphones.specout.com/d/l/Apple">iPhone</a>, <a href="https://companies.graphiq.com/l/8500602/Apple-Inc-in-Cupertino-CA">Apple Inc</a>.</p><p>Humans are quite effective at using contextual cues to properly understand questions. When confronted with ambiguity, the human mind works incredibly well at deciphering the correct intent. So in the case of the first example, you might say, “Yes, I’m an Apple fan. I’ve never had anything but an iPhone!”</p><p>Machines are not as gifted in this area, at least not today. Not only are machines bad at distinguishing the correct context of the content (e.g. fruit vs. company), but understanding related concepts is even harder (e.g. Gala or Granny Smith vs. iPhone or iPad). Deciphering the intent of an online searcher is inherently more challenging because that searcher is providing little to no context for the search engine (e.g. does the search query ‘best apple’ mean best apple fruit varieties or the best of Apple’s electronic products?).</p><p>This is where <a href="https://en.wikipedia.org/wiki/Semantic_search">semantic search</a> comes in. In short, semantic search attempts to augment and improve traditional search results by choosing the correct meaning and filtering out the noise to match the user’s intent.</p><h3><strong>Keyword Search vs. Semantic Search</strong></h3><p>Traditional search engine technology (also known as keyword search) is based almost entirely on the relationship between the searcher’s query and the occurrence of words in billions of web documents. Companies like Google used to rely primarily on this technique in their search algorithms to guess what the searcher was looking for and serve the most relevant web page at the top of the search results page.</p><blockquote>Humans are quite effective at using contextual cues to properly understand questions. When confronted with ambiguity, the human mind works incredibly well at deciphering the correct intent.</blockquote><p>Semantic search works very differently than keyword search. Rather than matching the query based on the words and phrases found on web pages, it matches the query to highly structured data in the form of entities, attributes, and relationships between data points. Instead of serving web pages that are <em>likely</em> to contain your answer, semantic search attempts to understand your question and <em>actually</em> <em>answer it.</em> This makes it easy for semantic search engines to show related information, since they know exactly what you’re looking for.</p><h4><em>Best Smartphone Under $400</em></h4><p>When you search ‘best smartphone under $400’ on a keyword search engine, the search engine will go through millions of web pages that contain any combination of the words ‘best,’ ‘smartphone,’ ‘under,’ and ‘$400’ and return the web pages that most likely answer your question.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1xNOoWo4yEHm2ayMlRTa9A.png" /><figcaption>The result of a query on Google.</figcaption></figure><p>When you search the same query on a semantic search engine, you get a very different result. Rather than showing content that the keyword search engine believes is broadly relevant to the original search query, semantic search engines aim to eliminate the less relevant content by semantically understanding the searcher’s intent. While keyword search engines may return links to different lists of ‘best smartphones under $400,’ semantic search engines are able to actually identify the best smartphone under $400 using a set of attributes. They can even show information relevant to that phone (e.g. the price of an iPhone over time or key features of the phone).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e5NRxXIpu0oweaHTMXEFIg.png" /><figcaption>The result of a query in Graphiq Search.</figcaption></figure><h3><strong>Our Approach</strong></h3><p>Let’s look at our own semantic search engine, <a href="https://www.graphiq.com/search">Graphiq Search</a>, as an example of what a semantic search engine is capable of returning for an abstract query like ‘<a href="https://www.graphiq.com/search/search?cid=24&amp;query=terrorism">terrorism</a>.’ Graphiq Search is capable of returning the most relevant results by recognizing the query’s connections to recent events, terror organizations, and locations.<strong> </strong>In the case of someone searching ‘terrorism,’ we’re able to display information such as most deadly terrorist groups, political candidate’s view on terrorism, and civilian deaths in the most recent terrorist attacks. The effectiveness of a semantic search engine comes from its ability to understand data relationships rooted in structured data.</p><p>At Graphiq, our ultimate goal is to build the world’s largest and most interconnected base of human knowledge. Over the years, we’ve been hyper-focused on developing a new way to answer people’s questions: one that strives to understand the user’s intent and put keywords in context, as opposed to just matching keywords. By doing so, we’re building a new way to deliver knowledge semantically, built upon the billions of entities and relationships that make up our knowledge graph. Through its ability to deliver an expanded range of relevant content, Graphiq provides the searcher with the opportunity to gain knowledge and learn more.</p><iframe src="https://cdn.embedly.com/widgets/media.html?url=https%3A%2F%2Fw.graphiq.com%2Fw%2Fjd2A9eaCgwR&amp;src=https%3A%2F%2Fw.graphiq.com%2Fw%2Fjd2A9eaCgwR&amp;type=text%2Fhtml&amp;key=d04bfffea46d4aeda930ec88cc64b87c&amp;schema=graphiq" width="600" height="454" frameborder="0" scrolling="no"><a href="https://medium.com/media/43096415f3ee9c67ae82795a66f7bcec/href">https://medium.com/media/43096415f3ee9c67ae82795a66f7bcec/href</a></iframe><h3><strong>Future of Search</strong></h3><p>Traditional keyword search engines are increasingly adopting a semantic paradigm into their core search products (e.g. Google’s Knowledge Graph). We believe the future of search is not just about quickly retrieving a large trove of information, but rather about gathering results that are more focused on the intent of the search. Ultimately, our aim is for you to spend less time searching and more time learning.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=71808207cfcc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/graphiq-engineering/what-is-semantic-search-71808207cfcc">What is Semantic Search?</a> was originally published in <a href="https://medium.com/graphiq-engineering">Graphiq Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>