A content recommendation method and content recommendation system

Abstract

The invention relates to the technical field of data analysis and processing, in particular to a content recommendation method and a content recommendation system used for recommending interested contents to a target user. The method comprises the steps of firstly performing learning training on historical content viewing sequences of users based on a continuous-bag-of-words model in natural language processing to obtain the continuous-bag-of-words model so as to obtain content vectors of contents; secondly obtaining similar contents of contents viewed by the target user according to the content vectors of the contents; thirdly calculating interest degrees of the target user to the similar contents; and finally extracting a preset number of the contents with the highest interest degree of the target user and recommending the contents to the target user. According to the method and the system, description information, attributes or tags of the contents and the users are not used, so that the situation that the robustness of an algorithm becomes poor due to deficiency of the information of the contents and the users is avoided; meanwhile, the calculation speed is much higher than that of collaborative filtering and content-based recommendation algorithms; and in addition, the contents are represented by equal-length vectors, so that various existing similarity algorithms can be adapted.

Description

technical field [0001] The present invention relates to the technical field of data analysis and processing, in particular to a content recommendation method and a content recommendation system for recommending interested content to target users. Background technique [0002] As people gradually step into the information age, the world today is in an environment of information explosion and is facing a severe problem of information excess. In 2011 alone, the global data volume reached 1.8ZB, equivalent to more than 200GB of data generated by each person in the world. This growth trend is still accelerating, and it is conservatively estimated that in the next few years, data will always maintain an annual growth rate of 50%. Today, users of platforms such as major e-commerce companies and video players generate massive amounts of data every day. Therefore, how to effectively use the data generated by users is an urgent problem for Internet companies today. At this time, the...

AI Technical Summary

Problems solved by technology

[0004] The disadvantages of video-based collaborative filtering are: ①The recommendation effect depends on the amount and accuracy of user historical preference data; ②User history and preferences are stored in a sparse matrix, and the calculation on the sparse matrix has obvious problems. Human error preferences have a greater impact on accuracy; ③ Due to the large number of users and videos, the calculation of the user-video matrix will be very large, and it will be difficult to implement real-time recommendation

The disadvantages of content-based related recommendations are: ①The description information of the video will be missing, resulting in the inability to extract video attributes; ②The extracted video feature minutes must be accurate and have a certain practical significance, otherwise it is difficult to ensure the relevance of the recommendation results

Images