Development of a Framework for Textual Entailment Recognition in Malayalam Language

Abstract

This thesis is about a framework proposed for textual entailment recognition in Malayalam language. Recognizing Textual Entailment (RTE) is newlineidentifying an inferential relationship between pairs of texts. Inference is newlineidentified from a sentence, called text/premise, for another sentence, called newlinea hypothesis. The text/premise entails a hypothesis if the hypothesis infers newlineits meaning from the text. The identification framework can recognize texts newlinein three forms: entailments, contradictions, and neutrality. These categories newlinescale to many more dimensions. RTE is also known as Natural Language newlineInference (NLI). newlineSemantics is the most prominent feature that conveys the meaning of a newlinesentence. NLI, also called Recognizing Textual Entailment, is an essential newlinesemantic feature for almost all Natural Language Processing applications, newlinesuch as question-answering systems, summarization, information retrieval, newlineand information extraction. The problem of textual entailment has two newlineforms. They are recognition of textual entailments and generation of textual newlineentailments. newlineThis thesis focuses on Recognizing Textual Entailment, which is useful in newlinemany NLP applications. newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced