A Blog Clustering Approach Based on Queried Keywords

In recent years, blogs were used between friends and family members. Nowadays, more and more bloggers are willing to share information with others. Therefore, it is desirable that related blogs can be connected. Clustering is often used for establishing connections between blogs. Full-text keywords retrieval process (FKRP) is generally used to obtain the keywords of posts, and the keywords were then used to cluster posts. Another approach is to use tags or categories chosen by bloggers for their posts. However, the overhead of FKRP is relatively high and the use of tagging is associated with synonymy related issues. More recently, there is an emerging concept called social tagging, which is about having web resources tagged by the general public together. Similar to this concept, we developed a blog connect platform that collects queried keywords used by the people for searching for a post. With the queried keywords, formal concept analysis (FCA) associated with agglomerative hierarchical clustering method is used to cluster posts. The experimental results showed that the accuracy of the proposed approach is better than those of other studies.