http://www.idonato.com/2008/07/23/seamless-voice-recognition-recruiters-get-ready/ All about what keeps me up at night Fri, 04 Feb 2011 16:49:15 -0700 hourly 1 http://wordpress.org/?v=3.1 http://www.idonato.com/2008/07/23/seamless-voice-recognition-recruiters-get-ready/comment-page-1/#comment-51 gutmach Fri, 19 Sep 2008 17:48:05 +0000 http://www.idonato.com/?p=74#comment-51 I agree that we will have the full life experiences digitally archived for at least some people (as Microsoft's Gordon Bell has already been doing on himself for years in the MyLifeBits project - http://research.microsoft.com/barc/mediapresence/mylifebits.aspx ). But I think you are partly missing the point about parsing the text. I know sites like Everyzing and Blinkx are doing pioneering work parsing video content for its text and making that searchable. But hopefully Web 3.0 will feature parsers that handle the life experience in other formats in addition to text. IMHO, the process of parsing such rich data into text inherently reduces its level of meaning. We need something that analyzes the data in context, and making it text first may not be the best solution. I can't guess what that might be, but a simpler current analogy may help illustrate the importance of this: Companies offering face recognition software that identifies fugitives within a crowd have a powerful offering, but they are analyzing digital data in very different way than text parsing. I agree that we will have the full life experiences digitally archived for at least some people (as Microsoft’s Gordon Bell has already been doing on himself for years in the MyLifeBits project – http://research.microsoft.com/barc/mediapresence/mylifebits.aspx ). But I think you are partly missing the point about parsing the text. I know sites like Everyzing and Blinkx are doing pioneering work parsing video content for its text and making that searchable. But hopefully Web 3.0 will feature parsers that handle the life experience in other formats in addition to text. IMHO, the process of parsing such rich data into text inherently reduces its level of meaning. We need something that analyzes the data in context, and making it text first may not be the best solution. I can’t guess what that might be, but a simpler current analogy may help illustrate the importance of this: Companies offering face recognition software that identifies fugitives within a crowd have a powerful offering, but they are analyzing digital data in very different way than text parsing.

]]> http://www.idonato.com/2008/07/23/seamless-voice-recognition-recruiters-get-ready/comment-page-1/#comment-42 cleomeo Fri, 25 Jul 2008 00:37:16 +0000 http://www.idonato.com/?p=74#comment-42 Great, interesting posting! (few spelling misses, but I'll overlook them because it was appealing) me Great, interesting posting! (few spelling misses, but I’ll overlook them because it was appealing)
me

]]>