The Method for the Unknown Word Classification
Natural Language Processing is a hard task. For the real Natural Language Processing, it is a necessary technique to process the unknown words. In this paper, we introduce the method for understanding the unknown words means. Many terms are newly created and we do not find these words in dictionary. Unknown words are generally occurred by reflecting the new phenomenon and technology. Hence, unknown words are dramatically created because of rapid changes in society. However, it is a hard task to define the meaning of all unknown words in dictionary. So, in this paper, we focus on how the machine understands the unknown words means. We propose a method to classify unknown words using the relevancy values between all nouns in the document and their TF values.
KeywordsNatural Language Processing Semantic Relation Pseudo Code Hard Task Unknown Word
Unable to display preview. Download preview PDF.
- 1.Ishikawa, H., Ito, A., Makino, S.: Unknown Word Processing Using Bunsetsu-automaton, 2nd class of Technical report of IEICE, LK-92-17, pp.1–8 (in Japanese) (1993)Google Scholar
- 2.Kamioka, T., Anzai, Y.: Syntactic Analysis of Sentences with unknown words by Abduction Mechanism (in Japanese). Journal of Artificial Intelligence 3, 627–638 (1988)Google Scholar
- 3.Kubomura, C., Sakurai, T., Kameda, H.: Evaluation of Algorithms for Unknown Word Acquisition, Technical report of IEICE, TL96-6, pp. 21–30 (in Japanese) (1996)Google Scholar
- 4.Scott, S., Matwin, S.: Text Classification using WordNet Hypernyms. In: The Proceeding of Workshop – Usage of WordNet in Natural Language Processing Systems, Montreal, Canada (1998)Google Scholar
- 5.Gelbukh, A., Sidorov, G., Guzman, A.: Use of a Weighted Topic Hierarchy for Document Classification. In: Matoušek, V., et al. (eds.) Text, Speech and Dialogue in Poc. 2nd Inter-national Workshop, Czech Republic. LNCS (LNAI), vol. 92, pp. 130–135. Springer, Heidelberg (1999)Google Scholar