Sunday, 22nd March 2009
« Blog Item
ItemBeware of placing too much reliance on automated sentiment analysis was the message of a recent LiveWire posting on Dow Jones, Standard & Poors and algorithmic trading (http://www.vivavip.com/go/e17065). Now sentiment analysis is a prominent feature of a new venture from the Financial Times, currently in beta test mode: Newssift. Instead of simply supporting keyword search, Newssift aggregates and annotates global business news and then searches data based on meaning and relationships, claiming to cut out the ‘commercial clutter’ in the process (http://digbig.com/4ymfh). ‘There are only a few search engines that employ relationship-based or semantic algorithms, and to date there is no other that accomplishes refinement using a business point of view,’ claims Robin Johnson, Chief Executive Officer, FT Search. Potential users are invited to try Newssift at http://www.newssift.com - meanwhile, though, some even more tightly focused personalization is going on at Time Inc, which is currently experimenting with a customized print or online magazine called Mine (http://www.timeinc.com/mine). Choosing from among five out of eight titles published by subsidiaries of Time Warner Inc, readers answer a short questionnaire asking, for example, whether they prefer sushi to pizza or would rather have dinner with Leonardo da Vinci or Socrates. This profile will enable editors to choose the stories that make up each fortnightly personalized issue (http://digbig.com/4ymfg). Dreamed up by car manufacturer Lexus, it’s an advertising initiative rather than an editorially driven one. But the Daily Me has been a twinkle in newspaper publishers’ eyes for years, so no doubt the experiment will be followed with close interest. But what about the ‘unknown unknowns’? Politicians and regulators are falling over themselves at the moment, for example, in their rush to acknowledge that they didn’t see the banking crisis coming. Could tightly targeted offerings like FT Newssift or Time’s Mine actually end up making it even more difficult for decision makers to see what they don’t expect?
« Blog Item
Item URL: https://web.jinfo.com/go/blog/66972
Printed: Wednesday, 26th February 2020
© 1997-2020 Jinfo Limited