<?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: Search Ranking Update</title>
	<atom:link href="http://blog.secondlife.com/2008/03/07/search-ranking-update/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/</link>
	<description>By Linden Lab</description>
	<pubDate>Wed, 07 Jan 2009 20:29:03 +0000</pubDate>
	<generator>http://wordpress.org/?v=MU</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Waiting&#8230; &#171; dreambits</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-596554</link>
		<dc:creator>Waiting&#8230; &#171; dreambits</dc:creator>
		<pubDate>Fri, 11 Apr 2008 10:06:56 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-596554</guid>
		<description>[...] Greta Umarov puts it on the Second life Blog &#8216;This thing (Second Life) will never be more than a curiosity [...]</description>
		<content:encoded><![CDATA[<p>[...] Greta Umarov puts it on the Second life Blog &#8216;This thing (Second Life) will never be more than a curiosity [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil Deakins</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589649</link>
		<dc:creator>Phil Deakins</dc:creator>
		<pubDate>Tue, 11 Mar 2008 13:44:44 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589649</guid>
		<description>@Ann (again)

When LL first announced the new search, I posted in this blog that business people would create a myriad of alts for the Picks/rankings. Whether they took notice of that, or they'd already realised it, there's a good possibility that, what is currently being called a defect, is actually intentional - by design. If all av's Picks counted as links, then many business owners would make use of it to move up the rankings, and it could become a case the those with the biggest 'armies' are the most successful, regardless of relevancy, which would cause many more complaints than now.

Off the top of my head, I can't think of a reason to change the way it is now. I can't think of a reason why the Picks of all avs should be used for rankings. The new search ranks on relevancy with the limited data that it has to work with, and those Picks links are a very powerful ranking factor. If people could push their places to the top by using an army of alts with Picks, it would devalue the results, imo. So I don't see the current limitations as a fault, or a defect, or a bug. I see it as beneficial to people who use the new search.</description>
		<content:encoded><![CDATA[<p>@Ann (again)</p>
<p>When LL first announced the new search, I posted in this blog that business people would create a myriad of alts for the Picks/rankings. Whether they took notice of that, or they&#8217;d already realised it, there&#8217;s a good possibility that, what is currently being called a defect, is actually intentional - by design. If all av&#8217;s Picks counted as links, then many business owners would make use of it to move up the rankings, and it could become a case the those with the biggest &#8216;armies&#8217; are the most successful, regardless of relevancy, which would cause many more complaints than now.</p>
<p>Off the top of my head, I can&#8217;t think of a reason to change the way it is now. I can&#8217;t think of a reason why the Picks of all avs should be used for rankings. The new search ranks on relevancy with the limited data that it has to work with, and those Picks links are a very powerful ranking factor. If people could push their places to the top by using an army of alts with Picks, it would devalue the results, imo. So I don&#8217;t see the current limitations as a fault, or a defect, or a bug. I see it as beneficial to people who use the new search.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil Deakins</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589645</link>
		<dc:creator>Phil Deakins</dc:creator>
		<pubDate>Tue, 11 Mar 2008 13:17:26 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589645</guid>
		<description>@94 Ann. Yes, I'm aware of those ideas, and also that the group creation one was shown not to be true. I haven't tested either of them personally (have you?), but I've personally tested many things - way before the current discussion started. It's not a new thing. It's not a 2 week old defect. It was noticed way before then, and was discussed in an SL forum thread (not the current thread), and it's suprising that people seem to think that it's a new find. It isn't.

Now what was that about trying again after doing some research? ;)</description>
		<content:encoded><![CDATA[<p>@94 Ann. Yes, I&#8217;m aware of those ideas, and also that the group creation one was shown not to be true. I haven&#8217;t tested either of them personally (have you?), but I&#8217;ve personally tested many things - way before the current discussion started. It&#8217;s not a new thing. It&#8217;s not a 2 week old defect. It was noticed way before then, and was discussed in an SL forum thread (not the current thread), and it&#8217;s suprising that people seem to think that it&#8217;s a new find. It isn&#8217;t.</p>
<p>Now what was that about trying again after doing some research? <img src='http://s.wordpress.com/wp-includes/images/smilies/icon_wink.gif' alt=';)' class='wp-smiley' /> </p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ann Otoole</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589594</link>
		<dc:creator>Ann Otoole</dc:creator>
		<pubDate>Tue, 11 Mar 2008 02:43:24 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589594</guid>
		<description>@94 - Phil, the NPIOF accounts that are showing have created a group and/or placed a classified ad. There is something wrong and i am afraid you have not done your homework. there is a defect open in jira on this. and this defect appears to be less than 2 weeks old.

try again after you do some research.</description>
		<content:encoded><![CDATA[<p>@94 - Phil, the NPIOF accounts that are showing have created a group and/or placed a classified ad. There is something wrong and i am afraid you have not done your homework. there is a defect open in jira on this. and this defect appears to be less than 2 weeks old.</p>
<p>try again after you do some research.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Rachel Darling</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589543</link>
		<dc:creator>Rachel Darling</dc:creator>
		<pubDate>Mon, 10 Mar 2008 19:17:28 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589543</guid>
		<description>First - I agree that at minimum you should be able to pick which parcel inherits the search reference; if that's by making it the last parcel, then fine.

In terms of Search in general, though -- as has been pointed out, the current implementation is never going to reliably reflect relevance, any more than Google does. Just too easy to "game" the system.

If you really want to help customers/searchers find relevant results, I'd suggest implementing a ranking system, and/or give unique visitors the ability to assign keywords to parcels they visit. You could then base relevance on the 10 most-frequently-used keywords that were assigned by visitors, and how often they were assigned. In terms of ranking, if people game the system with alts assigning non-relevant keywords, visitors have the option to "rank" them lower accordingly, and the ranking number will also show up in Search next to their listing.</description>
		<content:encoded><![CDATA[<p>First - I agree that at minimum you should be able to pick which parcel inherits the search reference; if that&#8217;s by making it the last parcel, then fine.</p>
<p>In terms of Search in general, though &#8212; as has been pointed out, the current implementation is never going to reliably reflect relevance, any more than Google does. Just too easy to &#8220;game&#8221; the system.</p>
<p>If you really want to help customers/searchers find relevant results, I&#8217;d suggest implementing a ranking system, and/or give unique visitors the ability to assign keywords to parcels they visit. You could then base relevance on the 10 most-frequently-used keywords that were assigned by visitors, and how often they were assigned. In terms of ranking, if people game the system with alts assigning non-relevant keywords, visitors have the option to &#8220;rank&#8221; them lower accordingly, and the ranking number will also show up in Search next to their listing.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil Deakins</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589534</link>
		<dc:creator>Phil Deakins</dc:creator>
		<pubDate>Mon, 10 Mar 2008 18:51:14 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589534</guid>
		<description>@93 Ann. That's just a rumour, Ann. Avs with no payment info do show in search. It was 'researched' not long ago, and we found them showing - not many, but it only needs one to put paid to the theory, and we found more than one. Today's 'news' is old news, and inaccurate.</description>
		<content:encoded><![CDATA[<p>@93 Ann. That&#8217;s just a rumour, Ann. Avs with no payment info do show in search. It was &#8216;researched&#8217; not long ago, and we found them showing - not many, but it only needs one to put paid to the theory, and we found more than one. Today&#8217;s &#8216;news&#8217; is old news, and inaccurate.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Ann Otoole</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589504</link>
		<dc:creator>Ann Otoole</dc:creator>
		<pubDate>Mon, 10 Mar 2008 17:08:59 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589504</guid>
		<description>check this out for the new search.. accounts with no payment info on file are not indexed by the new search. therefore it appears their top picks are also excluded from the new search. perhaps this is a step forward in making the new search less gamed by profile pick payola systems.</description>
		<content:encoded><![CDATA[<p>check this out for the new search.. accounts with no payment info on file are not indexed by the new search. therefore it appears their top picks are also excluded from the new search. perhaps this is a step forward in making the new search less gamed by profile pick payola systems.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil Deakins</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589483</link>
		<dc:creator>Phil Deakins</dc:creator>
		<pubDate>Mon, 10 Mar 2008 15:01:43 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589483</guid>
		<description>@91. The LL creation dates won't come into it after they've made it so that the largest plot wins, so cutting a plot to make sure that the largest one is the one you want to maintain will work.</description>
		<content:encoded><![CDATA[<p>@91. The LL creation dates won&#8217;t come into it after they&#8217;ve made it so that the largest plot wins, so cutting a plot to make sure that the largest one is the one you want to maintain will work.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: the bat</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589477</link>
		<dc:creator>the bat</dc:creator>
		<pubDate>Mon, 10 Mar 2008 13:06:28 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589477</guid>
		<description>Subdividing your larger plot wont work Phil , if it has an older date , 1 of the "new"parcels will inherit that older date , not matter how many subdivides you do , and when you join that one , which you cant identify as divides are random , to the rest , all will take the older date , exaqctly the same as if you'd not subdivided. 
Only Last or First Selected whichever is easiest to code , will solve this problem to everyone's satisfaction and make the process predictable.</description>
		<content:encoded><![CDATA[<p>Subdividing your larger plot wont work Phil , if it has an older date , 1 of the &#8220;new&#8221;parcels will inherit that older date , not matter how many subdivides you do , and when you join that one , which you cant identify as divides are random , to the rest , all will take the older date , exaqctly the same as if you&#8217;d not subdivided.<br />
Only Last or First Selected whichever is easiest to code , will solve this problem to everyone&#8217;s satisfaction and make the process predictable.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Phil Deakins</title>
		<link>http://blog.secondlife.com/2008/03/07/search-ranking-update/#comment-589473</link>
		<dc:creator>Phil Deakins</dc:creator>
		<pubDate>Mon, 10 Mar 2008 11:56:23 +0000</pubDate>
		<guid isPermaLink="false">http://lindenlab.wordpress.com/?p=1714#comment-589473</guid>
		<description>@87 Sarah. Traffic used in the search system is an issue with some people because of camping and traffic bots, which some people don't want in SL, but traffic doesn't affect the new All search much at all. 

The only way that I can see how traffic and the new search has affected businesses is many people using the new search instead of the Places tab, which would improve business for those who don't manipulate the traffic figures, and vice versa. It does require businesses to optimise their places for the new search though, but that's down to them. The new search is very much like web search engines. You don't usually get top rankings just because a website exists and is relevant. You usually need to do things to get top rankings.</description>
		<content:encoded><![CDATA[<p>@87 Sarah. Traffic used in the search system is an issue with some people because of camping and traffic bots, which some people don&#8217;t want in SL, but traffic doesn&#8217;t affect the new All search much at all. </p>
<p>The only way that I can see how traffic and the new search has affected businesses is many people using the new search instead of the Places tab, which would improve business for those who don&#8217;t manipulate the traffic figures, and vice versa. It does require businesses to optimise their places for the new search though, but that&#8217;s down to them. The new search is very much like web search engines. You don&#8217;t usually get top rankings just because a website exists and is relevant. You usually need to do things to get top rankings.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
