A Content-Oriented Framework for Online Discussion Analysis

Abstract

Mining and extracting quality knowledge from online discussions is significant for the industrial and marketing sector, as well as for e-commerce applications. Most of the existing techniques model a discussion as a social network of users represented by a user-based graph. In this paper, we propose a new framework for discussion analysis. It is based on message-based graphs where each vertex represents amessage object and each edge points out which message the specific node replies to. The edges can be weighted by the keywords that characterize the exchanged messages. This model allows a content-oriented representation of the discussion and it facilitates the identification of discussion chains. We compare the two representations (user-based and message-based graphs) and we analyze the different information that can be extracted from them. Our experiments with real data validate the proposed framework and show the additional information that can be extracted from a message-based graph.