Context and Lexicon based Gender Identification of Noun Phrase for Gujarati Text using Hybrid Approach
Published: 2019
Author(s) Name: Chetana Tailor, Bankim Patel |
Author(s) Affiliation:
Locked
Subscribed
Available for All
Abstract
To understand a language, analysis shall be performed at word level, sentence level, semantic level and discourse level. Morphological analysis is a foundation to perceive the meaning of a sentence. One of the essential tasks at morphological analysis is gender identification of a noun phrase that plays a pivotal role in Natural Language Processing Applications like Anaphora Resolution, Machine Translation and Text Summarization. It is a difficult task as there are special cases in Gujarati nouns that are contrasted with regular Gujarati linguistics rules for identifying the gender of the noun phrase. We have proposed solution for context based gender identification of noun phrase using Hybrid approach incorporating lexicon comprises of Gujarati nouns. Achieved accuracy of context based gender identification of noun phrase is 98.63% on development set and 88.62% on testing corpus.
Keywords: Natural Language Processing, Morph Analysis, Gujarati
View PDF