Tag-Based Intellisense Resource Sharing

The recommendation work focuses on resource sharing, users' interaction, similarity interest among people and dynamic changes of information, which can be represented by user behaviour. In order to take full advantage of user action, this paper proposes preference feature spatial and constructs IFS-Tree to locate all resources and users in space and analysis objects' relationship. The main contribution includes: (1) Giving the formula to calculate the similarity between any word to formation the preference feature data dictionary; (2) Giving the formula to calculate contribution values of resources and users according to users' behavior; (3) Proposing the construction of preference feature spatial and IFS-Tree to organize, locate and query all spatial objects; (4) Analysis and group all the objects, and applying the algorithm to the Wushu service network platform, the results show that searching any object or any group object can achieve millisecond respond, and the distribution of objects has the character of sparsity and long tail. (C) 2011 Published by Elsevier Ltd.