Abstract: Describes a personal information filtering technology with time factor. When building a user profile,the number of user interest catalog is determined and feature vectors built up according to user's behavior dynamically. It can well simulate the user's interest profile and also attaches time factor to the feature vector,taking into account both the short-term and long-term interests,tracing the user's interest shift. When filtering information,it computes the distance between a document and the user profile, then adjusts the final score according to the time factor. The system can study 267 documents and evaluate 402 documents per second. When the recall is 70%, the precision is 57%.