Research On Personalized Intelligent Information Retrieval System

As the popularity of Web application increases, the amount of information on the web grows.dramatically and the information on web is updated frequently. The problem users are faced with today is no longer the lack of useful information, but that of finding information pertinent to their personal needs. Traditional Information Retrieval (IR) technologies have satisfied people to some extent, but they are still suffering from low recall and precision problems. Most commercial search engines provide all-purpose service, which cannot satisfy any query from different background with different intention at different time. Based on the current situation, a Personalized Intelligent Information Retrieval System (PIIRS) is designed and implemented in this paper, which aims to improve the precision of retrieval.After studying the domestic and international current situation of the development of IR techniques broadly, the author has realized the deficiency of existing IR systems and the development trend of IR systems in the future. Concerning the deficiency of existing IR systems, research work on PIIRS has been done in this paper. We combine Agent technology of AI area, machine learning technology and clustering technology together with traditional IR technology in construction of the PIIRS.This paper provides overall design philosophy and architecture of this system; explains the methods used in implementing personalization and intelligence in details and describes all the key techniques and algorithms applied in the system thoroughly. The PIIRS learns user's preference by observing user behaviors during their interaction with the system. User's interests are used in information filtering process, thus the precision of the relevance documents presented to users improves greatly and the result documents are more relevant to users intention. Among all the modules, Agent module is the crucial one in this system, which isin charge of gaining user's preference, building and updating user's preference model, processing personalized information filtering according to user's preference model.In order to provide personalized and intelligent information retrieval service, user profile model is proposed in this paper to presenting user's preference. Terms database, which is clustered into several categories according to different interests topics is the key component of personalized realization. Reinforcement learning is used in autonomous updating and maintaining of the user profile model and clustering is used in personalized information filtering. The author describes all the methods and the algorithms used in the implementation in details. The results of the experiment show that those technologies lead to excellent performance. The research work that has been done in this paper has made good contribution to personalization and intelligence of information retrieval area.