KU: Word Sense Disambiguation by Substitution

Proceedings of the 4th International Workshop on Semantic Evaluations (SemEval-2007), pages 207–214, Prague, June 2007. c⃝2007 Association for Computational Linguistics KU: Word Sense Disambiguation by Substitution Deniz Yuret Koc¸ University Istanbul, Turkey dyuret@ku.edu.tr Abstract Data sparsity is one of the main factors that make word sense disambiguation (WSD) difficult. To overcome this problem we need to find effective ways to use resources other than sense labeled data. In this paper I de- scribe a WSD system that uses a statistical language model based on a large unanno- tated corpus. The model is used to evalu- ate the likelihood of various substitutes for a word in a given context. These likelihoods are then used to determine the best sense for the word in novel contexts. The resulting system participated in three tasks in the Se- mEval 2007 workshop. The WSD of prepo- sitions task proved to be challenging for the system, possibly illustrating some of its lim- itations: e.g. not all words have good sub- stitutes. The system achieved promising re- sults for the English lexical sample and En- glish lexical substitution tasks. 1 Introduction A typical word sense disambiguation system is trained on a corpus of manually sense tagged text. Machine learning algorithms are then employed to find the best sense for a word in a novel context by generalizing from the training examples. The training data is costly to generate and inter-annotator agreement is difficult to achieve. Thus there is very little training data available: the largest single cor- pus of sense tagged text, SemCor, has 41,497 sense tagged words. (Yuret, 2004) observed that approxi- mately half of the test instances do not match any of the contextual features learned from the training data for an all words disambiguation task. (Yarowsky and Florian, 2002) found that each successive doubling of the training data only leads to a 3-4% error reduc- tion within their experimental range. Humans do not seem to be cursed with an expo- nential training data requirement to become profi- cient with the use of a word. Dictionaries typically contain a definition and one or two examples of us- age for each sense. This seems to be sufficient for a human to use the word correctly in contexts that share no surface features with the dictionary exam- ples. The 108 waking seconds it takes a person to become proficient in a language does not seem suf- ficient to master all the words and their different senses. We need models that do not require large amounts of annotated text to perform WSD. What possible process can explain our proficiency without relying on a lot of labeled data? Let us look at a concrete example: The two most frequent senses of the word “board” according to WordNet 3.0 (Fell- baum, 1998) are the “committee” sense, and the “plank” sense. When we hear a sentence like “There was a board meeting”, it is immediately obvious that the first sense is intended. One hypothesis is that a common sense inference engine …

Content truncated to 500 words