The Study And Application Of Media Item Communications Analysis With Association Rules And Decision Tree Algorithm

With the advent of the era of big data, the explosive growth of online media data. Not only through the cinema screen media item come into the majority of the use-r’s eye, widespread on major online media, microblogging platform and the mainstre-am video sites.Item’s network communication influence consist of the amount of user watch the item on Video site, the amount of the item appear in aewspaper site, the amount of the item appear in microblogging. On the basis of summarize the data mining theory make it apply to in media item communications analysis. The thesis aims to finish the work as the following.First of all, count the amount of user watch the item on nearly ten video site, the amount of the item appear in nearly one hundred newspaper site, to find out the item which user watch most on video site or appear in newspaper site most, so as to prov-ide a basis suggestion for the advertising company advertising.Secondly, data mining the video site log data of users watch video on the video site with association rules algorithm. Then count the amount of user watch the item on nearly two thousand video site, find the association rules of ranking media item, to guide the video site’s purchase plan.Thirdly, data mining the films data of users watch films on the video site with decision tree algorithm. Find the relationship between film and show-times, movie categories, the main actors, directors and other properties. The experimental results show that where is a great relationship between movies showtimes attributes and the amount of users watch movies. And showtimes attributes become a major factor affe-ctting the amount.Finally, analysis of the linear relationship between the amount of user watch the item on video site and the amount of the item appear in microblogging or the amount of the positive item appear in sina microblogging just for "Where is the father" as an example. And make a detailed explanation of the results.This paper finds the hidden knowledge of item spread in the net with association rules and decision tree algorithm. The result model has a supporting role in real busi-ness decisions, and the results are satisfactory.