Extracting fine-grained durations for verbs from Twitter
Jennifer A. Williams · ePrints Soton (University of Southampton) · 2012
This paper presents recent work on a new method to automatically extract finegrained duration information for common verbs using a large corpus of Twitter tweets. Regular expressions were used to extract verbs and durations from each tweet in a corpus of more than 14 million tweets with 90.38 % precision covering 486 verb lemmas. Descriptive statistics for each verb lemma were found as well as the most typical fine-grained duration measure. Mean durations were compared with previous work by Gusev et al. (2011) and it was found that there is a small positive correlation. 1