Hierarchical Organization of Collaboratively Constructed Content

Abstract

AbstractHuge collections of collaboratively constructed content (e.g. blogs, consumer reviews, etc.) are now available online. This content has become a valuable knowledge repository, which enables users to seek quality information. However, such content is often unorganized, leading to difficulty in information navigation and knowledge acquisition. This chapter focuses on discovering the structure of the content and organizing them accordingly, so as to facilitate users in understanding the knowledge inherent within the content. In particular, we employ one example of the collaboratively constructed content, i.e. consumer reviews on products, as a case study, and propose a domain-assisted approach to generate a hierarchical structure to organize the reviews. The hierarchy organizes product aspects as nodes following their parent-child relations. For each aspect, the reviews and corresponding opinions on this aspect are stored. Such hierarchy provides a well-visualized way to browse consumer reviews at different granularity to meet various users’ needs, which can help to improve information dissemination and accessibility. We further apply the generated hierarchy to support the application of opinion Question Answering (opinion-QA) for products, which aims to generate appropriate answers for opinion questions about products. The experimental results on 11 popular products in 4 domains demonstrate the effectiveness of our approach.