Abstract:As a type of crowdsensing media, micro-blog has become an important crowdsensing place for a lot of real-time information dissemination and discussion. With the increasing of micro-blog users, there are more and more new topics emerging on this kind of platform, which has made the users difficult in finding out their own interesting topics. To solve this problem, this paper proposes a micro-blog topic recommendation system which can give corresponding suggestions/strategies for users. Firstly, the user relationship (i.e., a user adds a follow hyperlink to another user) in micro-blog can be effectively analyzed and saved to the user graph. In addition, an algorithm of computing user authority (which is similar to the idea of PageRank) is proposed to catch influential users based on the built user graph. Secondly, Topic Feature Graph (TFG) and User Micro-blog Feature Graph (UMFG) are respectively constructed based on the micro-blog text corpus of a topic and the micro-blog texts followed by a given user. Based on TFG and UMFG, User Topic Feature Vector (UTFV) and User Topic Feature Matrix (UTFM) can be achieved. After that, users’ similarity is calculated based on the User Topic Feature Vector and User Topic Feature Matrix to realize the users clustering by the help of the hierarchical clustering algorithm. Incorporating topic heat degree and user authority, the recommendation algorithm is presented to realize Micro-blog topic personalized recommendation within user clustering set. Experiments show that our proposed recommendation system has a good accuracy which is up to 50.2%.

Interest Mining and Recommendation System for Blog

Real-time Recommendation Method Based on Browsing Preferences Mining

A Content-Based Intelligent Ranking Model for Microblog

User Modeling for Recommendation in Blogspace

Personalized Microblogging Recommendation Based on Dynamic Interests and Social Networking of Users

An Online Blog Reading System by Topic Clustering and Personalized Ranking

SOCIAL INTEREST FOR USER SELECTING ITEMS IN RECOMMENDER SYSTEMS

Mining User Interests from Information Sharing Behaviors in Social Media.

Ranking Using Multi-Features In Blog Search

Mining the Personal Interests of Microbloggers via Exploiting Wikipedia Knowledge

Dynamic Semantic Clustering Approach For Web User Interest

Mining the Interests of Chinese Microbloggers Via Keyword Extraction

Mining Microblog User Interests Based on TextRank with TF-IDF Factor

DSIM: Dynamic and Static Interest Mining for Sequential Recommendation

Hybrid Recommendation Model Based on Social Media and Social Networks

Enhanced Semantic Graph Based Approach With Sentiment Analysis For User Interest Retrieval From Social Sites

The Recommendation System of Micro-Blog Topic Based on User Clustering

A System to Manage and Mine Microblogging Data.

Web-mining Algorithm for Recommendation System:Air

Mining Reviews for Product Comparison and Recommendation

Measuring and Visualizing Interest Similarity between Microblog Users