Talk to my Agents: Research Issues in Combining Information Filtering Agents with Collaborative Filtering

As the Internet increases our ability to access information, it also increases the severity of information overload -- the inability to find meaningful information among the slew of non-relevant information. Researchers have explored several methods for reducing information overload -- each with their strengths and their weaknesses. Information Filtering (IF) extracts item content and makes recommendations based on matches with a user interest profile. Collaborative Filtering (CF) matches users with other users with similar tastes to theirs, and makes recommendations based on the opinions of others in these "neighborhoods." This paper describes a series of experiments conducted by members of the GroupLens Research Project which show that a CF framework can be used to combine either generic filtering agents or personalized IF agents with the opinions of a community of users to produce better recommendations than either agents or users can produce alone. We have also shown that using CF to create a personal combination of a set of agents produces better results than either individual agents or other combination mechanisms. One key implication of these results is that users can avoid having to select among agents; they can use them all and let the CF framework select the best ones for them