Paper: Learning To Detect Conversation Focus Of Threaded Discussions

ACL ID N06-1027
Title Learning To Detect Conversation Focus Of Threaded Discussions
Venue Human Language Technologies
Session Main Conference
Year 2006
Authors

In this paper we present a novel feature- enriched approach that learns to detect the conversation focus of threaded discus- sions by combining NLP analysis and IR techniques. Using the graph-based algo- rithm HITS, we integrate different fea- tures such as lexical similarity, poster trustworthiness, and speech act analysis of human conversations with feature- oriented link generation functions. It is the first quantitative study to analyze hu- man conversation focus in the context of online discussions that takes into account heterogeneous sources of evidence. Ex- perimental results using a threaded dis- cussion corpus from an undergraduate class show that it achieves significant per- formance improvements compared with the baseline system.