Microsoft researchers are investigating if using data from several members of a social group, a technique Microsoft calls "groupization," can lead to better search results. Initial findings based on experiments involving about 100 participating Microsoft employees suggest that using different types of groups could produce significantly improved results. Microsoft researchers have developed an algorithm that, on average, pinpoints at least one search result for all members of a group that they judge to be better than other results returned by conventional search algorithms.
Microsoft says the new approach could help the company overcome an industry-wide plateau in the quality of search results. "Today, search engines are really challenged and are sort of at the cusp of having to know individuals better," says Microsoft Research computer scientist Jaime Teevan. The researchers are exploring how people with similar interests or attributes search for information, grouping people using explicit factors such as their age, gender, participation in certain mailing lists, and job function. In some cases, implicit groups, such as people who appeared to be performing the same task or have similar interests, were inferred. The researchers found that groups defined by demographics such as age and location have little in common for most searches, but groups of people with similar interests tend to rank similar search terms highly.
From Technology Review
View Full Article
No entries found