Predicting Thread Linking Structure by Lexical Chaining

Predicting Thread Linking Structure by Lexical Chaining Li Wang,♠♥ Diana McCarthy♦ and Timothy Baldwin♠♥ ♠ Dept. of Computer Science and Software Engineering, University of Melbourne ♥ NICTA Victoria Research Laboratory ♦ Lexical Computing Ltd li.wang.d@gmail.com, diana@dianamccarthy.co.uk, tb@ldwin.net Abstract Web user forums are valuable means for users to resolve specific information needs, both interactively for participants and stati- cally for users who search/browse over histor- ical thread data. However, the complex struc- ture of forum threads can make it difficult for users to extract relevant information. Thread linking structure has the potential to help tasks such as information retrieval (IR) and thread- ing visualisation of forums, thereby improv- ing information access. Unfortunately, thread linking structure is not always available in fo- rums. This paper proposes an unsupervised ap- proach to predict forum thread linking struc- ture using lexical chaining, a technique which identifies lists of related word tokens within a given discourse. Three lexical chaining algo- rithms, including one that only uses statistical associations between words, are experimented with. Preliminary experiments lead to results which surpass an informed baseline. 1 Introduction Web user forums (or simply “forums”) are online platforms for people to discuss and obtain informa- tion via a text-based threaded discourse, generally in a pre-determined domain (e.g. IT support or DSLR cameras). With the advent of Web 2.0, there has been rapid growth of web authorship in this area, and forums are now widely used in various areas such as customer support, community development, interactive reporting and online education. In ad- dition to providing the means to interactively par- ticipate in discussions or obtain/provide answers to questions, the vast volumes of data contained in fo- rums make them a valuable resource for “support sharing”, i.e. looking over records of past user inter- actions to potentially find an immediately applica- ble solution to a current problem. On the one hand, more and more answers to questions over a wide range of domains are becoming available on forums; on the other hand, it is becoming harder and harder to extract and access relevant information due to the sheer scale and diversity of the data. Previous research shows that the thread linking structure can be used to improve information re- trieval (IR) in forums, at both the post level (Xi et al., 2004; Seo et al., 2009) and thread level (Seo et al., 2009; Elsas and Carbonell, 2009). These inter- post links also have the potential to enhance thread- ing visualisation, thereby improving information ac- cess over complex threads. Unfortunately, linking information is not supported in many forums. While researchers have started to investigate the task of thread linking structure recovery (Kim et al., 2010; Wang et al., 2011b), most research efforts focus on supervised methods. To illustrate the task of thread linking recovery, we use an example thread, made up of 5 posts from 4 distinct participants, from the CNET forum dataset of Kim et al. (2010), as shown in Figure 1. The link- ing structure of the thread is modelled …

Content truncated to 500 words