<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" 
      xmlns:thr="http://purl.org/syndication/thread/1.0">
  <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php" />
  <link rel="self" type="application/atom+xml" href="http://www.readwriteweb.com/atom.xml" />
  <id>tag:,2008:/1/tag:www.readwriteweb.com,2008://1.5918-</id>
  <updated>2008-09-24T11:49:41Z</updated>
  <title>Comments for SemanticHacker Offers Cynical Bounty for Semantic Apps</title>
  
  <generator uri="http://www.sixapart.com/movabletype/">Movable Type 4.1</generator>
  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918</id>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php" />
    <link rel="service.edit" type="application/atom+xml" href="http://www.readwriteweb.com/cgi-bin/mt/mt-atom.cgi/weblog/blog_id=1/entry_id=5918" title="SemanticHacker Offers Cynical Bounty for Semantic Apps" />
    <published>2008-03-19T17:09:33Z</published>
    <updated>2008-03-19T18:06:59Z</updated>
    <title>SemanticHacker Offers Cynical Bounty for Semantic Apps</title>
    <summary>Contextual ad platform Textwise launched SemanticHacker.com today, a contest for applications and business plans built on the company&apos;s semantic analysis technology. While the new API is offered for free, the contest will award three winners with $100k and one winner could score up to $1 million based on subsequent commercialization. Semantic APIs only make sense,...</summary>
    <author>
      <name>Marshall Kirkpatrick</name>
      <uri>http://www.readwriteweb.com</uri>
    </author>
    
    <category term="Semantic Web" />
    
    <content type="html" xml:lang="en" xml:base="http://www.readwriteweb.com/">
      <![CDATA[<p><img src="http://www.readwriteweb.com/images/semhacklogo.jpg">Contextual ad platform Textwise launched <a href="http://SemanticHacker.com">SemanticHacker.com</a> today, a contest for applications and business plans built on the company's semantic analysis technology.  While the new API is offered for free, the contest will award three winners with $100k and one winner could score up to $1 million based on subsequent commercialization.</p>

<p>Semantic APIs only make sense, superior contextual advertising is clearly the most obvious cash-cow of semantics and a bounty sounds like a good idea - but there's something about this contest that doesn't quite feel right.</p>]]>
      <![CDATA[<p><img src="http://www.readwriteweb.com/images/semhackscreen.jpg" align="right" hspace="5px" vspace="5px"><h2>"One Meeelion Dollars"...A Cynical Gimmick?</h2>As is probably always the case, this bounty looks like a very cheap way to acquire some development that will pay off in a far larger sum.  Textwise has spent years developing their technology so perhaps that's fair.  </p>

<p>Taking the technology for a test run on SemanticHacker.com doesn't produce very exciting results, though.   You can try it yourself right on the front page of the site. While the super-hyped <a href="http://twine.com">Twine</a> at least attempts to add all kinds of value to the text it analyzes, the Textwise engine's results are clearly suited only to general, topical and contextual advertising.</p>

<p>The polite way to put this, as the company did in its press outreach, is as follows:<br />
<blockquote>The Textwise technology is focused on discovery as opposed to the “extraction” and metadata-based approaches of most semantic web technologies.  This is not the same as the “who/what/when/where/how” thread that is based around RDF, OWL and other similar ontologies and approaches.  No - this is a semantic discovery tool that can take in huge chunks of text and grok their essential "aboutness". The core technology (with origins in DARPA and intelligence agency stuff) is designed to decode the "DNA" of documents and be able to tell you what they're about.</blockquote></p>

<p>Like we said, suited far better for advertising than for anything else more interesting.  If semantic technology delivers nothing more than spy-agency-built ad networks then that's going to be a real tragedy.</p>

<p>Semantic web expert Paul Miller <a href="http://blogs.zdnet.com/semantic-web/?p=121">points out</a> that the contest's limitation to US participants is also a real loss of opportunity and says he'd be much more interested in a bounty not tied to any particular vendor's semantic API.</p>

<h2>Market Context</h2>

<p>Semantic markup is ready to hit the big time now that <a href="http://www.readwriteweb.com/archives/yahoo_supports_semantic_web.php">Yahoo! has announced that it will index the stuff</a>.  A long list of other startups are doing something similar to what TextWise is doing today.  <a href="http://www.hakia.com/">Hakia</a> started <a href="http://www.readwriteweb.com/archives/hakia_licenses_semantic_search.php">licensing its semantic search technology</a> yesterday.  We've <a href="http://www.readwriteweb.com/archives/dapper_funding_the_semantic_web.php">written before</a> about <a href="http://dapperads.com">Dapper's forthcoming semantic ad network</a>.</p>

<p>We've offered <a href="http://www.readwriteweb.com/archives/reuters_open_calais_apps_interview.php">extensive coverage of the Reuters Open Calais</a> semantic API extravaganza here, as well.</p>

<p>There's a lot of movement in this space; the market is downright crowded.  A search through our "top blogs in the semantic web" custom search engine (from the <a href="http://www.readwriteweb.com/archives/toolkit-08.php">RWW toolkit for 2008</a>) shows that none of the 60 blogs we follow on this space have ever written about TextWise or the <a href="http://www.semanticexchange.com/">SemanticExchange</a> network they participate in.  Maybe that shows the limitations of our list of sources, but maybe it says that TextWise just got a whole lot more publicity for a relatively small amount of money.</p>]]>
    </content>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49703</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49703" />
    <title>Comment from IdeaTagger on 2008-03-19</title>
    <author>
        <name>IdeaTagger</name>
        <uri>http://www.ideatagging.com</uri>
    </author>
    <content type="html" xml:lang="en" xml:base="http://www.ideatagging.com">
        <![CDATA[<p>Could you perhaps clear this up for me please Marshall?</p>

<p>Like I posted over on TechCrunch, I am a little confused about what semantic web technologies are meant to do. On the one hand it appears as though they are meant to magically create structure out of unstructured data. On the other hand, some articles including this one suggest that if the semantic web is to be any good, web publishers have to do a lot of the structuring themselves, i.e. marking up. </p>

<p>Oh and since I would really like to see an app like this built (and since I am in the UK and can't enter the competition), I will repeat one of my ideas for a semantic web app here for maximum exposure:</p>

<p>Idea: A “People in this post” photo widget for blogs or general websites. We see such photos in magazine articles all the time. It would automatically detect names in the post and suggest photos from a database (or from the general web) that the author can accept/reject. The author will of course be able to add new names and photos to the database. Each photo in the widget will link to a page on a destination site with links to other blog posts about that person. Over time a map of relationships can be built for each name in the database, i.e. people that they were mentioned in posts with and degrees of (post) separation between people. So if I was mentioned in a post with Mike and you and Mike were mentioned in another, then you and I are two degrees from each other. This could potentially have some interesting applications. </p>

<p>I don't even know if the above really counts as a semantic web app but it could be useful nonetheless.</p>]]>
    </content>
    <published>2008-03-19T18:57:47Z</published>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49706</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49706" />
    <title>Comment from Marshall Kirkpatrick on 2008-03-19</title>
    <author>
        <name>Marshall Kirkpatrick</name>
        <uri></uri>
    </author>
    <content type="html" xml:lang="en" xml:base="">
        <![CDATA[<p>IdeaTagger - I tried and have been told succeeded in offering a very clear explanation of the basic ideas behind the semantic web in the post about Yahoo! at <a href="http://www.readwriteweb.com/archives/yahoo_supports_semantic_web.php">http://www.readwriteweb.com/archives/yahoo_supports_semantic_web.php</a></p>

<p>Check that out.  See also our post on "top down semantic web" to answer your specific question. That's something there isn't consensus on.</p>]]>
    </content>
    <published>2008-03-19T19:02:21Z</published>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49708</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49708" />
    <title>Comment from Scott Lawton (Blogcosm) on 2008-03-19</title>
    <author>
        <name>Scott Lawton (Blogcosm)</name>
        <uri>http://blogcosm.com/</uri>
    </author>
    <content type="html" xml:lang="en" xml:base="http://blogcosm.com/">
        <![CDATA[<p>@IdeaTagger: My own view is that the Semantic Web is largely neutral about HOW the semantics get added.  What's interesting is all the cool things that can be done once data is tagged.  (People who are directly involved with SemWeb are welcome to disagree; I'm just an interested observer.)</p>]]>
    </content>
    <published>2008-03-19T19:13:29Z</published>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49709</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49709" />
    <title>Comment from Cesar H Castro Jr on 2008-03-19</title>
    <author>
        <name>Cesar H Castro Jr</name>
        <uri>http://www.standardsociety.com</uri>
    </author>
    <content type="html" xml:lang="en" xml:base="http://www.standardsociety.com">
        <![CDATA[<p>@IdeaTagger: Have you checked out <a href="http://www.polarrose.com/" rel="nofollow">PolarRose</a> it does something you have proposed.</p>

<p>By semantics, most mean microformats, which are ways individuals mark up computer friendly code so that when indexed by bots (i.e. search engines) certain elements can be graphed. Elements like contact cards, blogrolls, resumes, feeds, reviews, social networks, and others.</p>

<p>If you can imagine this would greatly change the way the 'Net is cultivated.</p>]]>
    </content>
    <published>2008-03-19T19:31:46Z</published>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49724</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49724" />
    <title>Comment from IdeaTagger on 2008-03-19</title>
    <author>
        <name>IdeaTagger</name>
        <uri>http://www.ideatagging.com</uri>
    </author>
    <content type="html" xml:lang="en" xml:base="http://www.ideatagging.com">
        <![CDATA[<p>Thanks All for your responses to my comment. They have certainly improved my understanding of what the semantic web is or could be. I think use cases (examples or actual implementations) of real life applications are what will get the message through to people.</p>

<p>On the "Who will do the markup" question posed in Marshall's other post linked to above, I think this is a huge issue and opportunity for someone to solve. I can imagine a plug-in for Wordpress/CMS that enables me to highlight a piece of text while composing a post or web page, right-click and select from a list of mark-up tags - person, place, event, tool, more and Other. If I select other, I can enter my own tag. If I select more, I can select from other people's defined tags. It should allow me to select more than one and should be clever enough to pre-select tags for words that it recognises. This could do for regular text what links (URLs) have done for web pages and more. Oh and I agree - the incentive will have to be SEO. Where do I sign-up?</p>

<p>@Cesar, Polar Rose seems to want to help you identify people in existing photos on web pages. My idea above is to auto-display photos in a widget of people referred to in the text of an article. With the technology they have though, they are probably only a step away from this.</p>]]>
    </content>
    <published>2008-03-19T21:14:46Z</published>
  </entry>

  <entry>
    <id>tag:www.readwriteweb.com,2008://1.5918-comment:49783</id>
    <thr:in-reply-to ref="tag:www.readwriteweb.com,2008://1.5918" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php"/>
    <link rel="alternate" type="text/html" href="http://www.readwriteweb.com/archives/semantichacker_offers_bounty.php#c49783" />
    <title>Comment from Connie Kenneally on 2008-03-20</title>
    <author>
        <name>Connie Kenneally</name>
        <uri>http://www.semantichacker.com</uri>
    </author>
    <content type="html" xml:lang="en" xml:base="http://www.semantichacker.com">
        <![CDATA[<p>Thanks for your comments, Marshall. I just wanted to clarify a few things:</p>

<p>1. The contest can actually award up to 3 winners with up to a million each.  All winners qualify for the full award:  $100K, job offer, funded development for the application through TextWise, and revenue share on a 50/50 basis up to $1M on the first year of release.  Of course, if we all end up making a lot of money out of a contributed idea, then everyone involved gets to benefit from that too. </p>

<p>       2.  While TextWise is one of the developers of seminal IP work in extraction (see our IP portfolio here: <a href="http://www.textwiselabs.com/company/assets.html)," rel="nofollow">http://www.textwiselabs.com/company/assets.html),</a> the technology launched in Semantic Hacker, as you point out, is not based on extraction.   </p>

<p>We wanted to delivery a <b>highly scalable</b> Semantic topic/concept based engine.  Based on the numerous inquiries we’ve had about our work from third parties, we thought it would be a plus to let some really bright guys access the discovery technology; the <i>“What’s it all about,”</i> (topics)  rather than the <i>“who, what , where, when, why”</i> involved in extraction.  </p>

<p>Keyword search is what most people online are familiar with today, but it’s painfully limited. Relationships between people/places/things is where some vendors are headed for tomorrow. Our goal was to bridge the gap with semantic topic scalability <b>today</b>. It is important to understand what things are about.  There are many steps in getting to the ideal semantic web, but no one company is going to have all of the answers. We’re hoping to contribute one of the stepping stones.</p>

<p>3.	 We are not at all convinced that the contextual Ad matching is the “best and most obvious cash cow” for this base technology. In our experience, contextual merchandising is the very tip of the iceberg for our semantic discovery technology. </p>

<p>We genuinely hope that we can attract some bright folks to work with us in realizing and sharing in the potential of this technology.</p>

<p>Connie Kenneally<br />
CEO<br />
TextWise<br />
</p>]]>
    </content>
    <published>2008-03-20T15:03:53Z</published>
  </entry>

</feed>