Newsgroup Clustering Based On User Behavior - A Recommendation Algebra
Jussi Karlgren · 1994
User models are a tool for guiding system behavior in interactive systems, and their utility and properties, desirable and undesirable, have been investigated in this context. There are several ways of utilizing information about the user that have not been implemented, however. In this paper a scheme for users to peek at other users' user models to extract information is proposed, in an information retrieval or information filtering domain. The material used for the study is a set of .newsrc files. Keywords Data extraction from user models; clustering. Background Compared to an ordinarily untidy bookcase computerized systems for information retrieval may not always be better. In a normal bookcase interesting documents may be found next to each other, and someone looking for a certain document may unexpectedly find other interesting documents in the vicinity. They are interesting because someone placed them there, and they are placed there because they have some relation to the orig...