The Structure of Polysemy : A Study of Multi-sense Words Based on WordNet

The Structure of Polysemy: A study of multi-sense words based on WordNet Jen-Yi Lin*, Chang-Hua Yang**, Shu-Chuan Tseng*, Chu-Ren Huang* *Institute of Linguistics, Preparatory Office **Chinese Knowledge Information Processing Group, Institute of Information Science Academia Sinica 128 Sec.2 Academy Rd., Nankang, Taipei, Taiwan R.O.C. {jenyi, chuanghwa}@hp.iis.sinica.edu.tw {tsengsc, churen}@gate.sinica.edu.tw Abstract The issues in polysemy with respect to the verbs in WordNet will be discussed in this paper. The hypernymy/hyponymy structure of the multiple senses is observed when we try to build a bilingual network for Chinese and English. There are several types of polysemic patterns and a co-hypernym may have the same word form as its subordinates. Fellbaum (2000) dubbed autotroponymy that the verbs linked by manner relation share the same verb form. However, her syntactic criteria seem not compatible to the hierarchies in WN. Either the criteria or the network should be reconducted. For most verbs in WN 1.7, polysemous relations are unlikely to extend over 3 levels of IS-A relation. Highly polysemous verbs are more complicated and may be involved in certain semantic structures. Semi-automatic sense grouping may be helpful for multimlinguital information retrieveal. 1. Introduction WordNet (WN), which is a large scale, domain-dependent semantic network of English words, provides a broad-coverage of lexical information. It represents a system of semantic relations among words, between words and synsets' , and between synsets themselves (Miller, 1990, 1995). Two features of the system are concept definitions and an inheritance hierarchy of concept types. Rather than using lexical entries only, the design-is based on linguistic theories about cognitive organization of natural languages. English nouns, verbs, adjectives, and adverbs are arranged to synsets that are in turn linked through semantic relations such as antonymy, hypernymy, etc. Like other conventional dictionaries and thesauri, WordNet also provides different meanings for one word. The lexicographic database represents a complex linguistic structure in which a word form may carry multiple senses. These word senses that are related in systematic ways build different synsets for each sense of a word. A word meaning then, is the pairing of a word form with a synset. However, WordNet's sense distinctions are more fine-grained than other machine-readable dictionaries, resulting in abundant polysemy and difficulty of computation (Kilgarriff 1997). In this paper, we will pay particular attention to the issues in polysemy with respect to the verbs in WordNet 1.7 and attempt to find a typical hypernymy/hyponymy structure for the multiple senses of a word form. In the following sections we will briefly overview the verb hierarchies in WN and illustrate the patterns of sense clusters. The types of autotroponymy will be discussed as well. 1 A set of synonyms refering to the same concept is called a synset. Members of synsets may be simple words or compounds. 320 Table 1. The relations found in WN for nouns and verbs Semantic Relation Noun Verb Similar relation Synonymy Opposite relation� . Antonymy� - Superordinate- Subordinate relation Hypernymy Troponymy Hyponymy Part-whole/manner relation Holonymy Entailment Meronymy Cause (From WordNet. Fellbaum 1998) 2. Hierarchical Organizations of …

Content truncated to 500 words