Comments on: A conversation with Jeff Jonas about connecting dots http://blog.jonudell.net/2009/01/05/a-conversation-with-jeff-jonas-about-connecting-dots/ Strategies for Internet citizens Mon, 13 Feb 2012 06:40:07 +0000 hourly 1 http://wordpress.com/ By: theChickenTest : Bryce Johnson » Data is Information. Information needs IAs. http://blog.jonudell.net/2009/01/05/a-conversation-with-jeff-jonas-about-connecting-dots/#comment-133283 Wed, 07 Apr 2010 00:58:52 +0000 http://blog.jonudell.net/?p=909#comment-133283 [...] few months ago I found a podcast where Jon Udell speaks with the Chief Scientist of IBM Entity Analytics, Jeff Jonas, who discusses a set of themes woven through his work, explored on his blog, and captured in a [...]

]]> By: Coast to Coast Bio Podcast » Blog Archive » Episode 9: One coast http://blog.jonudell.net/2009/01/05/a-conversation-with-jeff-jonas-about-connecting-dots/#comment-126504 Thu, 15 Jan 2009 20:43:07 +0000 http://blog.jonudell.net/?p=909#comment-126504 [...] Jon Udell interviews Jeff Jonas  on the interviews with innovators : A recommended listen from Deepak [...]

]]> By: esteban kolsky http://blog.jonudell.net/2009/01/05/a-conversation-with-jeff-jonas-about-connecting-dots/#comment-126338 Tue, 06 Jan 2009 09:32:16 +0000 http://blog.jonudell.net/?p=909#comment-126338 the issue here is to get past words and focus on their meaning – rather their content, context, and intent. it does not matter if it is john or jon as long as know it is the same person. and the only way that happens is by associating meta-tags and tags that uniquely identify a person regardless of the horrible work we do mispelling their names.

now, that is cool stuff…

]]> By: orcmid http://blog.jonudell.net/2009/01/05/a-conversation-with-jeff-jonas-about-connecting-dots/#comment-126323 Mon, 05 Jan 2009 17:35:47 +0000 http://blog.jonudell.net/?p=909#comment-126323 When I do an ego search (ok, not often but for amusement and to see what no-follows I need to add to push useless self-references off the top page), I am always amused by the “Did you mean to search for ‘orchid’?” and the sidebar ads for florists.

So, in annotating the web with semantic information, we have to do better than that, but it seems like a labor-intensive activity and one that is doomed when the wisdom of the crowd swamps the correct association.

Name confusion seems like an interesting problem in the absence of objects that represent the partitioning. There are a lot of D E Hamilton and Dennis E Hamilton entries in phone books around the country. Some of those listings are for me at different locations at different times. There have even been incorrect listings that were for me. And I receive an occassional call from someone looking for someone with a similar name or who just might be related to me.

We manage with this noise, although sticky identity confusion can be a problem (no-fly lists and misdirected liens and warrants come to mind). I think we cannot expect perfection. But how do we obtain the flexibility needed for corrections and improvement? I wonder how quickly that becomes the question. We’ll have to find out.

]]>