Design And Implementation Of Mobile Micro Video Community System Based On Recommendation Algorithm

With the exponential growth of data,in the face of huge data,users cannot accurately obtain the required information,resulting in users needing to spend more energy to filter out the information they need,resulting in "information overload" problems.Therefore,the recommendation system came into being,and the recommendation algorithm has also been greatly developed.Among them,the collaborative filtering algorithm is widely used with good performance,but it still has problems such as cold start,sparse score data features,and timeliness of user preferences,which will affect the recommendation effect.This article proposes an item collaborative filtering algorithm based on search terms and time factors to optimize the above problems and improve the recommendation effect.The algorithm adopts the cascade method idea of hybrid recommendation algorithm.First,the time factor and the average value of user ratings are smoothed into the item similarity calculation of the traditional item-based collaborative filtering algorithm to obtain a preliminary video recommendation set,followed by the preliminary video recommendation set The video description or movie name of the middle video is segmented,and the keywords in the video description are extracted based on the TF-IDF keyword feature extraction algorithm after smoothing the TF-IDF calculation formula to represent the corresponding video.By combining the user's search terms and browsing information records,a final Top N recommendation list is obtained,and the recommendation effect is improved.For the user's cold start problem,a Top N recommendation list is directly generated for new users based on the heat.The algorithm selects the corresponding algorithm to generate a recommendation list set by judging whether the user is a new user.In this paper,through experiments and verification,the recommendation effect of this recommendation algorithm is superior to the traditional item-based recommendation algorithm.In recent years,new media such as mobile micro video has developed rapidly,and has been welcomed at home and abroad.WeChat,as a phenomenon-level social APP,carries a large number of users,and the lightweight applet launched by the WeChat development team is popular because it is easy to use,convenient,and does not require mobile terminal memory.This article is based on the research of micro video and recommendation algorithm,relying on Yangchengtong public transportation community platform and WeChat Mini Program,combined with the collaborative filtering algorithm of items based on search terms and time factors,to develop a mobile micro video community applet to recommend to users Personalized micro video.This article first analyzes the needs of the applet system,then designs the overall architecture,specific modules and database of the system,divides the system into a mobile micro video community applet user module and a background administrator system module,and finally uses the relevant software technology to implement The module realizes the functions of user uploading micro videos,likes,comments,reports,scoring,searching and following,as well as the administrator's management function of the micro video community.This article has been tested and the system as a whole is operating normally,and the effect is as expected.