Information retrieval: data structures and algorithms
Information retrieval: data structures and algorithms
Mining Access Patterns Efficiently from Web Logs
PADKK '00 Proceedings of the 4th Pacific-Asia Conference on Knowledge Discovery and Data Mining, Current Issues and New Applications
Efficiently mining frequent trees in a forest
Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining
The Journal of Machine Learning Research
Modeling online reviews with multi-grain topic models
Proceedings of the 17th international conference on World Wide Web
Opinion integration through semi-supervised topic modeling
Proceedings of the 17th international conference on World Wide Web
Tag-based social interest discovery
Proceedings of the 17th international conference on World Wide Web
Hi-index | 0.00 |
Web 2.0 applications attract more and more people to express their opinions on the Web in various ways. However, the explosively increasing information in social web sites requires an effective mechanism to timely filter and summarize social common interest, and the moderator needs this mechanism as well to recommend the proper posts and guide public discussions. In this paper, we discuss the problem of recommending post in online communities: we firstly cluster the posts in groups based on their semantic relations, then filter the potential clusters by computing the cluster's support, and finally select the recommended posts as content representatives considering global and local support from each clusters. We compare different feature selections between tags, keywords and topics on cluster formation, and discuss their differences. The human judgement in our experiment shows that the recommendation based on marked tags is much more effective and concise than those on keywords and hidden topics.