Research On The Method Of Reading Personalized Recommendation Based On The Analysis Of Context

Nowadays, with the rapid development of network and economic society, customers tend to get lost in the flood of choice owing to the diversity of product and service, while the substantial growth of core date makes people spend much more time and cost obtaining the information that they want. Although proactive services of search technique could provide screening services for people getting information service, such as search engines, etc., but the popular pervasive service cannot satisfy the ever-growing needs of individuation, neither provide satisfied and suitable services for potential users at a specified time and space. In order to reduce the cost of search information and meet the demand of personalization, therefore, recommendation services was born at the right moment, which could bring convenience for people’s lives. Personalized recommendation has become one of the recommendation services’research trends.Based on the recommend-related theoretical framework, algorithm implementation and technological development in domestic and overseas, this article takes subjective consciousness into consideration in terms of both user experiences from the research field of ergonomics and decision theory and method, in addition, from the perspective of convenient and satisfaction, improve the method of personalized recommendation by introducing the scenario analysis theory. Firstly, Bayesian networks are applied to build context reasoning modes for network reading applications as the information that affects customer preferences. Secondly, we build customer preferences mode by confirming the weight coefficient among scenario information, basic user information and historical information, in the meanwhile, form a list of individual recommendations on the basis of revised counting method about user similarity. Finally, through analysis of numerical example and experimental results, the simulation confirms the effectiveness of personalized recommendation according to context analysis and quality improvement in recommendation. It has been improved that the methods and designs of applying context theory in the research on personalized recommendations accuracy than traditional collaborative filtering algorithm, and it indicates different current contexts of users will influence the selection decisions in system applications to a certain extent.The context personalized recommender systems do a comprehensive analysis of system performance, user psychology and environmental factors to achieve the integration of human-computer interaction, and also give a far-reaching significance and value to service providers of the website for improving the user experience, increasing user stickiness hits.