Automated Quality Estimation of Collaboratively Created Content

Abstract

User-generated content (UGC) has garnered much attention as it allows users to consume and produce content at a fast rate. Although such flexibility encourages contribution and knowledge-sharing, it also raises a question about the quality of the content. In this work, we present an automated method to estimate the quality of the collaboratively created content on Web2.0 platforms using the example of wikiHow. We define quality as a collection of independent feature groups, each focusing on a distinct quality aspect. We analyze these quality indicators' contribution to estimating the content's quality.