Research On Weibo Information Recommendation Model Based On User Portrait

From the theoretical level,there are a lot of researchers on the recommendation system to pay great attention,its application is gradually expanding.At the same time,the corresponding number of functional requirements are also significantly increased.For example,users are usually very concerned about both the real-time requirements and the accuracy requirements.At the present stage,multi-round model is used in both the process of user's requirement analysis and the process of user's requirement analysis.For the early machine,the use of single-round mode of advantage is more obvious and prominent.However,from the point of view of multi-turn pattern,we should pay more attention to the enhanced algorithm.The author takes the micro-blog network as a prerequisite,and establishes a user profile as a prerequisite for the proposed model.The reason for building such a model is to be able to understand the user's understanding of the characteristics of the project and personal preferences for a detailed and thorough study,not only that,but also in the scientific and reasonable inclusion of personalized elements.In addition,during the prediction of the score,the author has a highly targeted choice of the introduction of weight factors such a means,which makes the prediction more efficient and error-free.If we compare the proposed model with the early recommendation Algorithm,it is not difficult to find that the former results are more accurate and the error is smaller.However,generally speaking,the model proposed by the author still has further improvement and improvement,and hopes to be further improved in the future study.