Research on Personalized Recommendation System of Scientific and Technological Periodical Based on Automatic Summarization

Abstract

Utilizing the theoretical methods and technology of automatic summarization system and personalized recommend system, ire study how to access to thesis document index, theme words, summary, readers' evaluations and other important recommended information from the vast amount of scientific and technological periodical documents quickly and effectively. The aim is to improve the scientific workers' research efficiency remarkably. On the basis of the traditional automatic summarization methods, we proposed the frame model based on automatic summarization system of the scientific and technological periodical; Utilizing the conceptual model of key technology of e-commerce recommendation system, ire discussed the main key technology of the recommendation system; based on the specific needs of the researchers to the recommendation system, ire probed into the Marking principle of the system thoroughly, and designed the personalized recommendation system frame. The main innovation is the recommendation system which has multi-level, flexible and controllable characteristics, and reflects a people-oriented, reader-centered thinking. Besides, the recommendation result is an integrated knowledge set document including thesis document index, theme words, summary and readers' evaluations etc.