Research On Context Awareness Based Hybird Recommendation Algorithm For WeChat Public Paltform

In recent years, with the rapid development of smart terminals and mobile Intemet,social applications on mobile terminal emerge in large numbers every day. WeChat was rolled out by Tencent in 2011, and the number of WeChat users has exceeded 650 million this year. WeChat supports more than 20 kinds of language and has been used over 200 countries, thus becoming a veritable super application on mobile terminal. As a significant part of WeChat, WeChat Public Platform changes the way people getting information. This article mainly analyses hot con-tent, self features and introduction information of WeChat Official Account, design a content recommendation system for WeChat Official Account,Which can provide highly-related and up-to-date information for a given WeChat Official Account, thus assisting the WeChat Official Account in publishing high-level content and attract more users'subscription.This article use hybrid recommendation algorithm based on content-based recommendation algorithm and collaborative filtering recommendation algorithm to provide WeChat Official Account precise results, which can assist WeChat Official Account in gaining user's traffic. The core of this recommendation system is mainly based on two parts. On the one hand, the system use content-based recommendation algorithm and use crawler to get all contents related to the appointed WeChat Official Account. Then we extract key words of all hot articles, count the key words frequency and vectorize WeChat Official Account text con-tent. On the other hand, the system use collaborative filtering recommendation algorithm to analyse the hot contents, self features and introduction information of a given WeChat Official Account, and extract other WeChat Official Account that are similar with the appointed WeChat Official Account from database. Then the system provides the hot contents from the extracted WeChat Official Account to the appointed Official Account.The content-based recommendation algorithm and collaborative filtering recommendation algorithm are all classical recommendation algorithm, but each single recommendation algorithm has its own shortcoming. So we often use hybrid recommendation algorithms in practical applications, in other words,we usually use more than two recommendation algorithms in practical appli-cations. This article use the hybrid recommendation algorithm based on content-based recommendation algorithm and collaborative filtering recommendation algo-rithm, and improve the collaborative filtering recommendation algorithm.By comparing the similarity between recommended results and hot contents, the simulation experiments shows that the recommendation system meets the requirements.In addition, the experiment proves that the unproved algorithm works well in this recommendation system.