Paper: The Good the Bad and the Unknown: Morphosyllabic Sentiment Tagging of Unseen Words

ACL ID P08-2028
Title The Good the Bad and the Unknown: Morphosyllabic Sentiment Tagging of Unseen Words
Venue Annual Meeting of the Association of Computational Linguistics
Session Main Conference
Year 2008
Authors

The omnipresence of unknown words is a problem that any NLP component needs to ad- dress in some form. While there exist many established techniques for dealing with un- known words in the realm of POS-tagging, for example, guessing unknown words’ semantic properties is a less-explored area with greater challenges. In this paper, we study the seman- tic field of sentiment and propose five methods for assigning prior sentiment polarities to un- known words based on known sentiment carri- ers. Tested on 2000 cases, the methods mirror human judgements closely in three- and two- way polarity classification tasks, and reach ac- curacies above 63% and 81%, respectively.