This paper improves traditional collaborative filtering algorithm,proposes a new user profile based on concept hierarchy tree,which can make recommender systems still work even when users have no common rating items.

It’s basic idea is to be construct the concept level tree by the database and the Web log,accords to express the itemsets support of the crossover level by application the interval support, and gives to mining multilevel association rule by application the method of the level tree pruning and the increment renewal.

Method of conception hierarchy tree construction based on qualitative mapping;

The algorithm of generating concept hierarchy of digital field;

New concept hierarchy optimization method in attribute-oriented induction;

Contrast research of Chinese domain ontology concept hierarchy induction methods

Research on Data Mining Algorithm Based on Conception Hierarchy Tree and Its Application;

A high-speed mining algorithm for multiple-leveled association rules is imposed in this paper, which uses samples to mining from a cer-taim intermediate level of the conception hierarchy, to increase speed of mining.

When primitive databases accessed are either undesirable or unavailable, this model can return approximate answers using generalized databases and concept hierarchies stored at mobile units.