Analysis of user critical interests based on an improved Shortest Root Tree algorithms

The resources of users and merchandise are vital part in business networks. The relationship that among user and merchandise gives a reflection of user interest and its habits that user tends to select some merchandise. The analysis of user critical interests plays an important role in E-business recommendation system. The paper introduces a shortest-path algorithm. It looks at user´s attribute as a root and user´s interests as a series of nodes. It puts forwards an Improved Shortest Root Tree algorithm to calculate path lengths which from root node to each objective node. Path lengths are used to analyze the most interested merchandise nodes of user. The experimentation shows that the algorithm can well describe user´s interest and it has both nicer performances in precision and less time costs.