<?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"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: LinkedIn Search Results Sorting: Relevance or Keyword?</title>
	<atom:link href="http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/</link>
	<description>Leveraging social networks, resume databases, and the Internet for sourcing and recruiting</description>
	<lastBuildDate>Sat, 13 Mar 2010 14:21:36 -0500</lastBuildDate>
	<generator>http://wordpress.org/?v=2.8.3</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: LinkedIn&#8217;s Relevance vs. Keyword Conundrum: How Do You Sort? &#124; Recruitalicious</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-5059</link>
		<dc:creator>LinkedIn&#8217;s Relevance vs. Keyword Conundrum: How Do You Sort? &#124; Recruitalicious</dc:creator>
		<pubDate>Tue, 24 Nov 2009 16:38:43 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-5059</guid>
		<description>[...] while I thought this was a great way to help me find the best candidates, after reading a recent Boolean Black Belt blog post, I realized that I&#8217;ve ignored a large part of what LinkedIn has to [...]</description>
		<content:encoded><![CDATA[<p>[...] while I thought this was a great way to help me find the best candidates, after reading a recent Boolean Black Belt blog post, I realized that I&#8217;ve ignored a large part of what LinkedIn has to [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jake Mannix</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-5035</link>
		<dc:creator>Jake Mannix</dc:creator>
		<pubDate>Wed, 18 Nov 2009 21:26:21 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-5035</guid>
		<description>Hi Glen,

  I could say a lot on this subject, but one thing I should mention is that you effectively describe LinkedIn&#039;s &quot;relevance&quot; sort as somehow &quot;different&quot; than Google&#039;s, but that&#039;s only true in the details: when you do a search on Google (or Yahoo, etc), do you think they only take into account the keywords and boolean structure you put into the query?  Of course not - they do fantastic work to limit spam and properly weight sites which are &quot;authorities&quot; on a subject (that&#039;s the whole magic of PageRank).

  LinkedIn&#039;s relevance sort is doing something similar, but since the data set isn&#039;t the web, PageRank doesn&#039;t make sense to help inform the keyword relevance.  What does make sense, to help filter out the noise, the mostly empty profiles, the people who maybe haven&#039;t updated in ages or are only connected to a couple friends and don&#039;t really have a presence on there, is to let the social graph act as a weighting filter in a similar way that the web-link graph acts for Google.

  The thing that&#039;s different for LinkedIn, in comparison to Google, is that instead of just using the global structure of the graph, the local structure (who *you* are connected to) can take part.  This helps you actually find a way to contact the people you&#039;ve found, once you find them (because they&#039;re more likely to be someone within two to three degrees away from you).

  Just like any good search engine, there is some domain-specific special sauce that goes into LinkedIn&#039;s search, because once you start looking at the data, you see that relevance is not such a cut-and-dry &quot;how many times did my search term match in the title vs. the description field?&quot;.

  -jake</description>
		<content:encoded><![CDATA[<p>Hi Glen,</p>
<p>  I could say a lot on this subject, but one thing I should mention is that you effectively describe LinkedIn&#8217;s &#8220;relevance&#8221; sort as somehow &#8220;different&#8221; than Google&#8217;s, but that&#8217;s only true in the details: when you do a search on Google (or Yahoo, etc), do you think they only take into account the keywords and boolean structure you put into the query?  Of course not &#8211; they do fantastic work to limit spam and properly weight sites which are &#8220;authorities&#8221; on a subject (that&#8217;s the whole magic of PageRank).</p>
<p>  LinkedIn&#8217;s relevance sort is doing something similar, but since the data set isn&#8217;t the web, PageRank doesn&#8217;t make sense to help inform the keyword relevance.  What does make sense, to help filter out the noise, the mostly empty profiles, the people who maybe haven&#8217;t updated in ages or are only connected to a couple friends and don&#8217;t really have a presence on there, is to let the social graph act as a weighting filter in a similar way that the web-link graph acts for Google.</p>
<p>  The thing that&#8217;s different for LinkedIn, in comparison to Google, is that instead of just using the global structure of the graph, the local structure (who *you* are connected to) can take part.  This helps you actually find a way to contact the people you&#8217;ve found, once you find them (because they&#8217;re more likely to be someone within two to three degrees away from you).</p>
<p>  Just like any good search engine, there is some domain-specific special sauce that goes into LinkedIn&#8217;s search, because once you start looking at the data, you see that relevance is not such a cut-and-dry &#8220;how many times did my search term match in the title vs. the description field?&#8221;.</p>
<p>  -jake</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Boolean Black Belt</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4970</link>
		<dc:creator>Boolean Black Belt</dc:creator>
		<pubDate>Wed, 04 Nov 2009 01:47:47 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4970</guid>
		<description>Ivan,
You are correct - the titles in recommendations are the same as in the experience section. What I find interesting is that they effectively count twice if they are positive hits as search terms, provided you have a recommendation for each position. And when it comes to &quot;relevance&quot; - having a bunch of title-hits doesn&#039;t impress me, nor does it have anything to do with true relevance for my search - at least not IMO. 

Your thoughts?</description>
		<content:encoded><![CDATA[<p>Ivan,<br />
You are correct &#8211; the titles in recommendations are the same as in the experience section. What I find interesting is that they effectively count twice if they are positive hits as search terms, provided you have a recommendation for each position. And when it comes to &#8220;relevance&#8221; &#8211; having a bunch of title-hits doesn&#8217;t impress me, nor does it have anything to do with true relevance for my search &#8211; at least not IMO. </p>
<p>Your thoughts?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ivan &#124; SEO Consultant.ie</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4903</link>
		<dc:creator>Ivan &#124; SEO Consultant.ie</dc:creator>
		<pubDate>Mon, 02 Nov 2009 15:00:52 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4903</guid>
		<description>Aren’t job Titles the same as the Titles for recommendations on the LinkedIN profile?

Or did I completely misunderstood you?</description>
		<content:encoded><![CDATA[<p>Aren’t job Titles the same as the Titles for recommendations on the LinkedIN profile?</p>
<p>Or did I completely misunderstood you?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Boolean Black Belt</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4866</link>
		<dc:creator>Boolean Black Belt</dc:creator>
		<pubDate>Sun, 01 Nov 2009 22:52:07 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4866</guid>
		<description>Eric,
Thank you for your comment! I agree 100% - this is an important question, and I am frankly surprised that this topic and line of questioning hasn&#039;t garnered more attention. 

As far as my testing and experimentation (albeit limited to this point) has shown so far, it appears that LinkedIn heavily favors keywords mentioned in title fields - including title headers for recommendations. Also, while I see keywords highlighted in the &quot;specialties&quot; section, they aren&#039;t highlighted in the &quot;summary&quot; section even when present - which seems odd to me. 

When I run searches and sort by keyword instead of LinkedIn&#039;s proprietary &quot;relevance&quot; sorting, I am not terribly impressed with the top ranked results - many have multiple title hits, but don&#039;t even have their experience sections filled in with any information. For me, true relevance has less to do with titles and more to do with responsibilities.

For job seekers, figuring out LinkedIn SEO is absolutely critical - because if sourcers and recruiters are running searches to find people with certain experience, it helps to know which fields are the most heavily weighed in how LinkedIn determines keyword relevance/sorting - because showing up at the BOTTOM of 100&#039;s of results is a sure way NOT to be found!</description>
		<content:encoded><![CDATA[<p>Eric,<br />
Thank you for your comment! I agree 100% &#8211; this is an important question, and I am frankly surprised that this topic and line of questioning hasn&#8217;t garnered more attention. </p>
<p>As far as my testing and experimentation (albeit limited to this point) has shown so far, it appears that LinkedIn heavily favors keywords mentioned in title fields &#8211; including title headers for recommendations. Also, while I see keywords highlighted in the &#8220;specialties&#8221; section, they aren&#8217;t highlighted in the &#8220;summary&#8221; section even when present &#8211; which seems odd to me. </p>
<p>When I run searches and sort by keyword instead of LinkedIn&#8217;s proprietary &#8220;relevance&#8221; sorting, I am not terribly impressed with the top ranked results &#8211; many have multiple title hits, but don&#8217;t even have their experience sections filled in with any information. For me, true relevance has less to do with titles and more to do with responsibilities.</p>
<p>For job seekers, figuring out LinkedIn SEO is absolutely critical &#8211; because if sourcers and recruiters are running searches to find people with certain experience, it helps to know which fields are the most heavily weighed in how LinkedIn determines keyword relevance/sorting &#8211; because showing up at the BOTTOM of 100&#8217;s of results is a sure way NOT to be found!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Eric Jaquith</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4700</link>
		<dc:creator>Eric Jaquith</dc:creator>
		<pubDate>Thu, 29 Oct 2009 20:22:11 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4700</guid>
		<description>Glen,

This is a VERY important question that needs to be discussed. Many LinkedIn users and job seeker are not aware that the default relevant sort  results that hiring managers and recruiters are seeing, may make them irrelevant and therefore invisible. The average LinkedIn user does not use complex keyword Boolean search. They enter a job title or a company and select a zip code or city. They review the top 10 or 20 result and rarely look at the top 100 available to the standard user.

Most of us get great &quot;relevant&quot; results with Google, so most searchers just trust LinkedIn search is like Google. What does a job seeker need to do to be in the top 10? Can someone game the system and really not be relevant? Are bad candidate doing this now?

Sincerely,

Eric Jaquith

Keep up the great work!


PS: Sorting by connections is only available in the advance search beta and was removed from the standard advance search. Users need to go to the people advance search page and look below the search button at the bottom. They will see this&quot;

 &quot; New! Try our new People Search Beta featuring dynamic search refinements. &quot;

 Click this link access the beta screen
http://www.linkedin.com/search?optIn=</description>
		<content:encoded><![CDATA[<p>Glen,</p>
<p>This is a VERY important question that needs to be discussed. Many LinkedIn users and job seeker are not aware that the default relevant sort  results that hiring managers and recruiters are seeing, may make them irrelevant and therefore invisible. The average LinkedIn user does not use complex keyword Boolean search. They enter a job title or a company and select a zip code or city. They review the top 10 or 20 result and rarely look at the top 100 available to the standard user.</p>
<p>Most of us get great &#8220;relevant&#8221; results with Google, so most searchers just trust LinkedIn search is like Google. What does a job seeker need to do to be in the top 10? Can someone game the system and really not be relevant? Are bad candidate doing this now?</p>
<p>Sincerely,</p>
<p>Eric Jaquith</p>
<p>Keep up the great work!</p>
<p>PS: Sorting by connections is only available in the advance search beta and was removed from the standard advance search. Users need to go to the people advance search page and look below the search button at the bottom. They will see this&#8221;</p>
<p> &#8221; New! Try our new People Search Beta featuring dynamic search refinements. &#8221;</p>
<p> Click this link access the beta screen<br />
<a href="http://www.linkedin.com/search?optIn=" rel="nofollow">http://www.linkedin.com/search?optIn=</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Claire</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4569</link>
		<dc:creator>Claire</dc:creator>
		<pubDate>Tue, 27 Oct 2009 10:54:51 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4569</guid>
		<description>Hi Glen,
I certainly hadn&#039;t realisd how Linked In were defining &#039;relevance&#039;, but actually I&#039;m finding it works in my favour. As I&#039;m sourcing for an energy company, when I did a very loose search on tax professionals based in Norway, at the top of the list were tax people who had worked in energy firms. I&#039;m not sure how it was calculated, if it&#039;s down to the fact I have a lot of energy professionals in my network, or I&#039;m down as working in an energy company, or a member of a number of energy related groups, but either way it works for me!
However, now I&#039;m aware that no matter how carefully I craft my search strings, that may not bring forward the best skills matches first, I&#039;ll rethink how I view the results in future.

Thanks again!</description>
		<content:encoded><![CDATA[<p>Hi Glen,<br />
I certainly hadn&#8217;t realisd how Linked In were defining &#8216;relevance&#8217;, but actually I&#8217;m finding it works in my favour. As I&#8217;m sourcing for an energy company, when I did a very loose search on tax professionals based in Norway, at the top of the list were tax people who had worked in energy firms. I&#8217;m not sure how it was calculated, if it&#8217;s down to the fact I have a lot of energy professionals in my network, or I&#8217;m down as working in an energy company, or a member of a number of energy related groups, but either way it works for me!<br />
However, now I&#8217;m aware that no matter how carefully I craft my search strings, that may not bring forward the best skills matches first, I&#8217;ll rethink how I view the results in future.</p>
<p>Thanks again!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Daniel Alfon</title>
		<link>http://www.booleanblackbelt.com/2009/10/linkedin-search-results-sorting-relevance-or-keyword/comment-page-1/#comment-4510</link>
		<dc:creator>Daniel Alfon</dc:creator>
		<pubDate>Mon, 26 Oct 2009 17:15:14 +0000</pubDate>
		<guid isPermaLink="false">http://www.booleanblackbelt.com/?p=4290#comment-4510</guid>
		<description>Glen,

Thanks for this interesting post. The fact that &quot;Search people&quot; is such a frequent action makes this post actionable even to LinkedIn users who only have a white belt.  

You may want to know searching by Connection is still not available to all users. 

I am going to stress the Keywords Vs. Relevance point you made in my free presentation trying to help out job seekers with 50-300 connections - thank you... http://www.slideshare.net/danielalfon/advanced-linkedin-guide-for-puzzled-2009-job-seekers

All the best,</description>
		<content:encoded><![CDATA[<p>Glen,</p>
<p>Thanks for this interesting post. The fact that &#8220;Search people&#8221; is such a frequent action makes this post actionable even to LinkedIn users who only have a white belt.  </p>
<p>You may want to know searching by Connection is still not available to all users. </p>
<p>I am going to stress the Keywords Vs. Relevance point you made in my free presentation trying to help out job seekers with 50-300 connections &#8211; thank you&#8230; <a href="http://www.slideshare.net/danielalfon/advanced-linkedin-guide-for-puzzled-2009-job-seekers" rel="nofollow">http://www.slideshare.net/danielalfon/advanced-linkedin-guide-for-puzzled-2009-job-seekers</a></p>
<p>All the best,</p>
]]></content:encoded>
	</item>
</channel>
</rss>
