Analysis of Japanese Compound Nouns by Direct Text Scanning

Abstract

This paper aims to analyze word dependency structure in compound nouns appearing in Japanese newspaper articles. The analysis is a dil't:icult problem because such compound nouns can be quite long, have no word boundaries between contained nouns, and often contain nnregistered words such as abbreviations. The nonsegmentation property and unregistered words cause initial segmentation errors which result in erroneous analysis. This paper presents a corpus-based approach which scans a corpus with a set of pattern matchers and gathers cooccurrence examples to analyze compound nouns. It employs boot-strapping search to cope with unregistered words: if an unregistered word is lound in the process of searching the examples, it is recorded and invokes additional searches to gather the examples containing it.

Experimental evaluation of ranking and selection methods in term extraction

“…To extract promising term candidates of compound noun and at the same time to exclude undesirable strings such as is a or of the, the most frequently used method is to filter out the words being members of the stop-list. In these days, more complex structures like noun phrases, collocations consisting of nouns, verbs, prepositions, determiners, and so on, become focused on (Smadja and McKeown 1990:252-259;Ananiadou 1996:41-46, Zhai andEvans 1996:17-23;Hisamitsu andNitta 1996:550-555, Shimohata et al 1997:476-481). All of these are good term candidates in a document or a specific domain because all of them have a strong unithood.…”

Resolving Ambiguous Segmentation of Korean Compound Nouns Using Statistics and Rules

“…To analyze the structure of compound nouns in Japanese, one of several methods uses collocation information and a thesaurus Tanaka 1994, 1995). Another method is a corpus-based approach that scans a corpus with a set of pattern matchers and gathers co-occurrence examples to analyze compound nouns (Hisamitsu and Nitta 1996). For analyzing compound nouns in English, one method compares the adjacency model and the dependency model and concludes that the dependency model provides a substantial advantage over the adjacency model (Lauer 1995).…”