<?xml version="1.0" encoding="UTF-8" standalone="yes"?><oembed><version><![CDATA[1.0]]></version><provider_name><![CDATA[The Dish]]></provider_name><provider_url><![CDATA[http://dish.andrewsullivan.com]]></provider_url><author_name><![CDATA[Andrew Sullivan]]></author_name><author_url><![CDATA[https://dish.andrewsullivan.com/author/sullydish/]]></author_url><title><![CDATA[Yeah, Secret Service, Like That&#8217;s Gonna&nbsp;Happen]]></title><type><![CDATA[link]]></type><html><![CDATA[<p>The agency is <a href="http://www.washingtonpost.com/blogs/the-fix/wp/2014/06/03/the-secret-service-wants-software-that-detects-social-media-sarcasm-yeah-sure-it-will-work/" target="_blank">looking</a> for an intrepid software developer to create a sarcasm detector for social media. Mary Beth Quirk <a href="http://consumerist.com/2014/06/04/you-could-be-the-person-who-builds-sarcasm-detection-software-for-the-secret-service-no-really/">sums it up</a> thus:</p>
<blockquote><p>Basically, the Secret Service would love it if someone would explain the Internet so it doesn’t go around arresting sarcastic people with itchy social media trigger fingers.</p>
<p>Another thing that sounds a bit weird?</p></blockquote>
<p><!--tpmore --></p>
<blockquote><p>The software will have the “functionality to send notifications to users.” Because that wouldn’t freak someone out to get a popup window from the Secret Service just being like, “Hey, did you mean that like, for real? Or are you being sarcastic? Thanks, juuust checking in!”</p></blockquote>
<p>But Jesse Singal <a href="http://nymag.com/scienceofus/2014/06/computers-are-still-terrible-at-getting-sarcasm.html">doubts</a> they&#8217;ll come up with anything:</p>
<blockquote><p><a href="http://aclweb.org/anthology//P/P11/P11-2102.pdf">One study from 2011</a> (PDF) used tweets that had been specifically hashtagged #sarcasm or #sarcastic, stripped those hashtags, and then dumped them into a virtual pile with a bunch of other straightforwardly positive and negative tweets. At their best performance, the computer programs the researchers used could only correctly separate sarcastic from non-sarcastic tweets about 65 percent of the time — and this was in a rather controlled setting.</p>
<p><a href="http://www.cs.uic.edu/~liub/">Bing Liu</a>, a University of Illinois at Chicago computer scientist who authored a <a href="http://www.amazon.com/Sentiment-Analysis-Opinion-Mining-Bing-ebook/dp/B009KET3PU">book</a> about sentiment analysis (that is, extracting emotional context from text), expressed skepticism that anyone yet has a good handle on this problem. &#8220;I am not aware that anyone has a satisfactory algorithm or system that can detect sarcastic sentences,&#8221; he said in an email. And the stuff the Secret Service would be looking at would be a particularly uphill battle: &#8220;In discussions about politics [sarcasm] is fairly common and very hard to deal with because it often requires some background knowledge which computers are not good at.&#8221;</p></blockquote>
]]></html></oembed>