CFILT: Resource Conscious Approaches for All-Words Domain Specific WSD

Proceedings of the 5th International Workshop on Semantic Evaluation, ACL 2010, pages 421–426, Uppsala, Sweden, 15-16 July 2010. c⃝2010 Association for Computational Linguistics CFILT: Resource Conscious Approaches for All-Words Domain Specific WSD Anup Kulkarni Mitesh M. Khapra Saurabh Sohoney Pushpak Bhattacharyya Department of Computer Science and Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India {anup,miteshk,saurabhsohoney,pb}@cse.iitb.ac.in Abstract We describe two approaches for All-words Word Sense Disambiguation on a Spe- cific Domain. The first approach is a knowledge based approach which extracts domain-specific largest connected com- ponents from the Wordnet graph by ex- ploiting the semantic relations between all candidate synsets appearing in a domain- specific untagged corpus. Given a test word, disambiguation is performed by considering only those candidate synsets that belong to the top-k largest connected components. The second approach is a weakly super- vised approach which relies on the “One Sense Per Domain” heuristic and uses a few hand labeled examples for the most frequently appearing words in the target domain. Once the most frequent words have been disambiguated they can pro- vide strong clues for disambiguating other words in the sentence using an iterative disambiguation algorithm. Our weakly supervised system gave the best perfor- mance across all systems that participated in the task even when it used as few as 100 hand labeled examples from the target do- main. 1 Introduction Domain specific WSD exhibits high level of ac- curacy even for the all-words scenario (Khapra et al., 2010) - provided training and testing are on the same domain. However, the effort of creating the training corpus - annotated sense marked corpora - for every domain of interest has always been a matter of concern. Therefore, attempts have been made to develop unsupervised (McCarthy et al., 2007; Koeling et al., 2005) and knowledge based techniques (Agirre et al., 2009) for WSD which do not need sense marked corpora. However, such approaches have not proved effective, since they typically do not perform better than the Wordnet first sense baseline accuracy in the all-words sce- nario. Motivated by the desire to develop annotation- lean all-words domain specific techniques for WSD we propose two resource conscious ap- proaches. The first approach is a knowledge based approach which focuses on retaining only domain specific synsets in the Wordnet using a two step pruning process. In the first step, the Wordnet graph is restricted to only those synsets which contain words appearing in an untagged domain- specific corpus. In the second step, the graph is pruned further by retaining only the largest con- nected components of the pruned graph. Each tar- get word in a given sentence is then disambiguated using an iterative disambiguation process by con- sidering only those candidate synsets which ap- pear in the top-k largest connected components. Our knowledge based approach performed better than current state of the art knowledge based ap- proach (Agirre et al., 2009). Also, the precision was better than the Wordnet first sense baseline even though the F-score was slightly lower than the …

Content truncated to 500 words