Abstract
The aim of this paper is to quantitatively investigate the sorts of expressions that are used around the boundaries of conversation topics in multi-party meeting conversations. This will be done by means of identifying word collocations which are statistically-significantly associated with topic boundaries, and graphically presenting them in the form of a network. The International Computer Science Institute Meeting Recorder Dialogue
Act Corpus is used in this study. We will demonstrate that the derived network of collocated words supports previous studies undertaken in the area of conversation analysis on topic changes in some respects, while other findings of conversation analysis are not confirmed by the empirical results of our study. This study presents expressions which
seem to be distinctive to multi-party meeting conversations by referring to the nature of meeting conversations
Original language | English |
---|---|
Publication status | Published - 2010 |
Event | Conference of the Australian Linguistic Society 2009 - Melbourne Australia Duration: 1 Jan 2010 → … |
Conference
Conference | Conference of the Australian Linguistic Society 2009 |
---|---|
Period | 1/01/10 → … |