Who is helping Whom?

Who is helping Whom?

There have been successful efforts to produce a "personalized search" with some niche search engines, such as Eurekster (www.eurekster.com). For example, two friends might share a common interest in a subject, such as Star Trek the original series, so one of the two goes to Eurekster and puts in the "trekie" search term. He finds what he/she wants but on the second page so he/she opens it up to view the material. Now the second friend who also has the same interest in the same material, goes to Eurekster to find it. Since the former friend has already located it the latter friend will now find that the site is at the top of the list on the first page with a little Eurekster "e" icon next to it. That's to alert them that their friend in their network likes the same site. Since both friends share a common interest, the search will probably feel more relevant to each one. However Eurekster assumes you have contacted "friends" of similar interests and then formed a "search group" among you. Eurekster "remembers the names, their choices and the members of the search group.

The latest technology attempts to create a collaborative grouping. "This technology looks for patterns in people's likes and dislikes, and uses those patterns to help people find things they did not know they were looking for. Computer Sciences calls this "finding good things". Collaborative filtering also has the power to do the converse, 'keep bad things away', for instance by filtering unsolicited commercial e-mail messages, or spam.

Finding unknown good things, however, can at present only be done using collaborative filtering." Economist Mar 10 05

Collaborative filtering builds its data base on individual preferences (previously collected), then compiles a large group (peer) of people with similar preferences (unbeknownst to them) and uses the characteristics of the group to pre-select other items/services the individual might be interested in. As the data banks enlarge the technology will be able to match more individuals to their respective peer groups pre-selecting likely individual preferences.