Paper: Acquisition Of Kowledge Data By Analyzing Natural Language

ACL ID C86-1106
Title Acquisition Of Kowledge Data By Analyzing Natural Language
Venue International Conference on Computational Linguistics
Session Main Conference
Year 1986

Thus, after examing 78,000 four character kanji strings, 28,000 different kinds of knowledge data was obtained from 32,000 types of kanji strings in net. This approach is promising because it allows easy expansion in quantity and consists mainly of mechanical processing. In addition, an attempt to reorganize a total of 887,000 data, 200,000 in net, is under way. An atten~t to apply the same idea to the three and five character kanji strings is also being made. 449 {Example] T~/~ Pv @:~. ~ T~Pv ~• ~J~ T ga/o Pv kaicho-jinin TniyoriPv shaahin-hantei 4-2 Aquisition of knowledge data based on word-toword synchronism In this ease, Japanese sentences should be analyzed to extract connections and obrain knowledge data. Analysis, however, takes a great deal of time and involves ambiguity. This s...