Semantic network discussion representation: applicability and some potential benefits

Collaborative tools typically employ linear threading for representing and organizing the contributions of the discussants. We argue that linear representations provide little support for the conceptualization, contextualization, and visualization of the discussion. To address these drawbacks, we propose the usage of semantic networks for representing discussions. In this paper, we demonstrate the applicability of semantic network representation to discussions through the development of a prototype named CollectiveThought. We also test some of the potential benefits of semantic network discussion representation with an empirical study. More specifically, we compare these two modes of discussion representation (i.e., linear versus semantic network) in terms of effects on the communication process as measured by perceived contextualization and on the communication outcome as measured by mutual understanding, using both subjective and objective indicators. The results show that semantic network discussion representation leads to a higher level of perceived contextualization and better mutual understanding.