<?xml version="1.0" encoding="utf-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Enterprise search still a technology conversation</title>
	<atom:link href="http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation/feed" rel="self" type="application/rss+xml" />
	<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation</link>
	<description>Peter Van Dijck's weblog</description>
	<pubDate>Wed, 03 Dec 2008 02:47:57 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.5.1</generator>
		<item>
		<title>By: Nitin Kewalramani</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-102902</link>
		<dc:creator>Nitin Kewalramani</dc:creator>
		<pubDate>Mon, 27 Nov 2006 09:20:17 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-102902</guid>
		<description>Hi Peter,
I came across your name and your expertise on google. I am a recruitment executive based in London working for Computer People and I a am looking for someone who is a speacialist in Verity K2 Enterprise Edition 5.5. The position is based in Geneva, Switzerland. If you are interested please reply to this message. Thanks</description>
		<content:encoded><![CDATA[<p>Hi Peter,<br />
I came across your name and your expertise on google. I am a recruitment executive based in London working for Computer People and I a am looking for someone who is a speacialist in Verity K2 Enterprise Edition 5.5. The position is based in Geneva, Switzerland. If you are interested please reply to this message. Thanks</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Bob Leach</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-19362</link>
		<dc:creator>Bob Leach</dc:creator>
		<pubDate>Mon, 20 Jun 2005 13:46:14 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-19362</guid>
		<description>"&lt;i&gt;Some robots are less foolish than others, but no robot is as wise as a human editor&lt;/i&gt;." 

This is quote I like from Richard Wiggins, a Michigan State University information technologist responsible for the creation of MSU Keywords, a non-spidered implementation of Best Bets for the MSU websites. He writes about it in "Beyond the Spider - The Accidental Thesaurus" (http://www.infotoday.com/searcher/oct02/wiggins.htm). To illustrate the value of such an approach, he says to "visit your favorite university on the Web and search for 'map.' Most people doing that search want a campus map. In many cases the spider will offer high on the hit list the library's map of Mesopotamia before the university's map of its own campus. Thanks to MSU Keywords, we deliver the campus map as the first item on the hit list."</description>
		<content:encoded><![CDATA[<p>&#8220;<i>Some robots are less foolish than others, but no robot is as wise as a human editor</i>.&#8221; </p>
<p>This is quote I like from Richard Wiggins, a Michigan State University information technologist responsible for the creation of MSU Keywords, a non-spidered implementation of Best Bets for the MSU websites. He writes about it in &#8220;Beyond the Spider - The Accidental Thesaurus&#8221; (http://www.infotoday.com/searcher/oct02/wiggins.htm). To illustrate the value of such an approach, he says to &#8220;visit your favorite university on the Web and search for &#8216;map.&#8217; Most people doing that search want a campus map. In many cases the spider will offer high on the hit list the library&#8217;s map of Mesopotamia before the university&#8217;s map of its own campus. Thanks to MSU Keywords, we deliver the campus map as the first item on the hit list.&#8221;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ben Anderson</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-17229</link>
		<dc:creator>Ben Anderson</dc:creator>
		<pubDate>Thu, 09 Jun 2005 17:39:02 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-17229</guid>
		<description>I agree with all this here. What Jeff Lsah is talking about is graphically displayed by Tanya Rabourn's graph (the line is a bit faint but the course it plots is pretty remarkable) of search queries at: http://www.pixelcharmer.com/fieldnotes/archives/process_designing/2003/000326.html . Also if an organisation has discussion boards you'll probably see the same questions come up time and again - e.g. for a large government organisation you get things like: where are this year's payscales on the intranet? What is the hottest/coldest our offices are supposed to be? etc. Having Best Bets as out-of-the-box functionality on intranet search just makes so much sense. Maintenance of correct results is an added cost of course...</description>
		<content:encoded><![CDATA[<p>I agree with all this here. What Jeff Lsah is talking about is graphically displayed by Tanya Rabourn&#8217;s graph (the line is a bit faint but the course it plots is pretty remarkable) of search queries at: <a href="http://www.pixelcharmer.com/fieldnotes/archives/process_designing/2003/000326.html" rel="nofollow">http://www.pixelcharmer.com/fieldnotes/archives/process_designing/2003/000326.html</a> . Also if an organisation has discussion boards you&#8217;ll probably see the same questions come up time and again - e.g. for a large government organisation you get things like: where are this year&#8217;s payscales on the intranet? What is the hottest/coldest our offices are supposed to be? etc. Having Best Bets as out-of-the-box functionality on intranet search just makes so much sense. Maintenance of correct results is an added cost of course&#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Walter Underwood</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16255</link>
		<dc:creator>Walter Underwood</dc:creator>
		<pubDate>Thu, 26 May 2005 16:39:55 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16255</guid>
		<description>In the Verity Ultraseek product, they are called Quick Links, and have been in the product for a while, since late 2001. We also added a feature to mine the clickthrough data and suggest new Quick Links to be added.

If hand-selected matches are shown in-line with algorithmic results, we call that "editorial results" or "result stuffing". Traditionally, Best Bets are a separate set of hits, clearly called out as editorial.

I think the term Best Bets was first used by Infoseek, but the sponsored keyword matches at AOL were probably earlier.

I like your other recommendations, and we'll add them to our list of neat things to do.</description>
		<content:encoded><![CDATA[<p>In the Verity Ultraseek product, they are called Quick Links, and have been in the product for a while, since late 2001. We also added a feature to mine the clickthrough data and suggest new Quick Links to be added.</p>
<p>If hand-selected matches are shown in-line with algorithmic results, we call that &#8220;editorial results&#8221; or &#8220;result stuffing&#8221;. Traditionally, Best Bets are a separate set of hits, clearly called out as editorial.</p>
<p>I think the term Best Bets was first used by Infoseek, but the sponsored keyword matches at AOL were probably earlier.</p>
<p>I like your other recommendations, and we&#8217;ll add them to our list of neat things to do.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Madonnalisa</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16254</link>
		<dc:creator>Madonnalisa</dc:creator>
		<pubDate>Thu, 26 May 2005 15:21:01 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16254</guid>
		<description>I actually have a document where I compared several enterprise search engines that included some kind of best bets functionality.  We ended up with Verity's Search that was formerly Inktomi Enterprise Search.  I believe it was called Recommended Links.  Definitely a great topic to bang away at, to really get the vendors learning more is to get the customers to demand the functionality.  I believe that's how HP did it with Inktomi's Ultraseek.</description>
		<content:encoded><![CDATA[<p>I actually have a document where I compared several enterprise search engines that included some kind of best bets functionality.  We ended up with Verity&#8217;s Search that was formerly Inktomi Enterprise Search.  I believe it was called Recommended Links.  Definitely a great topic to bang away at, to really get the vendors learning more is to get the customers to demand the functionality.  I believe that&#8217;s how HP did it with Inktomi&#8217;s Ultraseek.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16253</link>
		<dc:creator>Phil</dc:creator>
		<pubDate>Thu, 26 May 2005 10:25:29 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16253</guid>
		<description>&lt;i&gt;Here is an incomplete list of products that do best bets, and what they call it.&lt;/i&gt;
[...]
&lt;i&gt;
    * Verity: yes, calls them Sponsored Links.
&lt;/i&gt;
Hmmm?

&lt;i&gt;By the way, to work well with users, best bets should appear in-line with the other search results, not separate from them.&lt;/i&gt;

I'm not familiar with Verity &#038; may be misreading. But the day Google put their Sponsored Links inline will be a good day for A8.</description>
		<content:encoded><![CDATA[<p><i>Here is an incomplete list of products that do best bets, and what they call it.</i><br />
[...]<br />
<i><br />
    * Verity: yes, calls them Sponsored Links.<br />
</i><br />
Hmmm?</p>
<p><i>By the way, to work well with users, best bets should appear in-line with the other search results, not separate from them.</i></p>
<p>I&#8217;m not familiar with Verity &#038; may be misreading. But the day Google put their Sponsored Links inline will be a good day for A8.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kevin Crossman</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16252</link>
		<dc:creator>Kevin Crossman</dc:creator>
		<pubDate>Wed, 25 May 2005 19:38:44 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16252</guid>
		<description>Google's Search Appliance has this functionality.</description>
		<content:encoded><![CDATA[<p>Google&#8217;s Search Appliance has this functionality.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Gammel</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16251</link>
		<dc:creator>David Gammel</dc:creator>
		<pubDate>Wed, 25 May 2005 02:41:26 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16251</guid>
		<description>Wow, sounds like they don't get down in the trenches very often. 

BTW, Ultraseek does best bets as well, although that may be what you meant by Verity.</description>
		<content:encoded><![CDATA[<p>Wow, sounds like they don&#8217;t get down in the trenches very often. </p>
<p>BTW, Ultraseek does best bets as well, although that may be what you meant by Verity.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jeff Lash</title>
		<link>http://poorbuthappy.com/ease/archives/2005/05/20/2683/enterprise-search-still-a-technology-conversation#comment-16170</link>
		<dc:creator>Jeff Lash</dc:creator>
		<pubDate>Tue, 24 May 2005 01:14:44 +0000</pubDate>
		<guid isPermaLink="false">/?p=2683#comment-16170</guid>
		<description>Just wanted to second your post -- best bets are important, and unfortunately not as easy to implement or as common as they should be for how important they are. I've been doing a fair amount of search log analysis on some sites lately, and the top 100/200 unique queries make up a fairly significant percentage of the overall search queries. I'm trying to advocate for a best bets sort of approach -- take the top 100 queries, and make sure the best results show up in the page (or, even better, make sure the entire first page of results is good -- but many people still seem to be of the opinion that "the engine will take care of it" if tuned correctly. Did I miss something -- are computers suddenly smarter than humans?</description>
		<content:encoded><![CDATA[<p>Just wanted to second your post &#8212; best bets are important, and unfortunately not as easy to implement or as common as they should be for how important they are. I&#8217;ve been doing a fair amount of search log analysis on some sites lately, and the top 100/200 unique queries make up a fairly significant percentage of the overall search queries. I&#8217;m trying to advocate for a best bets sort of approach &#8212; take the top 100 queries, and make sure the best results show up in the page (or, even better, make sure the entire first page of results is good &#8212; but many people still seem to be of the opinion that &#8220;the engine will take care of it&#8221; if tuned correctly. Did I miss something &#8212; are computers suddenly smarter than humans?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
