Corpus-based metrics for assessing communal common ground


This article presents the first attempt to construct a computational model of common ground. Four corpus-based metrics are presented that estimate what facts are likely to be in common ground. The proposed metrics were evaluated in an experiment with human participants, focussing on a domain of famous people. The results are encouraging: two of the proposed metrics achieved a large positive correlation between the estimates of how widely known a property of a famous person is and the percentage of participants who knew the corresponding property.

Back to Table of Contents