PROJECT TITLE :

Finding Related Forum Posts through Content Similarity over Intention-based Segmentation - 2017

ABSTRACT:

We have a tendency to study the problem of finding related forum posts to a post at hand. In distinction to traditional approaches for finding connected documents that perform content comparisons across the content of the posts as a full, we have a tendency to think about each post as a collection of segments, each written with a completely different goal in mind. We advocate that the relatedness between two posts should be primarily based on the similarity of their respective segments that are intended for the identical goal, i.e., are conveying the same intention. This means that it is attainable for the identical terms to weigh differently within the relatedness score relying on the intention of the segment in which they are found. We have a tendency to have developed a segmentation methodology that by monitoring a number of text options can determine the parts of a post where vital jumps occur indicating a purpose where a segmentation should occur. The generated segments of all the posts are clustered to make intention clusters and then similarities across the posts are calculated through similarities across segments with the same intention. We tend to experimentally illustrate the effectiveness and efficiency of our segmentation methodology and our overall approach of finding related forum posts.