APPLIED TEXTUAL ENTAILMENT

This book introduces the applied notion of textual§entailment as a generic empirical task that captures§major semantic inferences across many applications.§Textual Entailment addresses semantic inference as a§direct mapping between language expressions and§abstracts the common semantic inferences as needed§for text based Natural Language Processing§applications. The book defines the task and describes§the creation of a benchmark dataset for textual§entailment along with proposed evaluation measures.§It further describes how textual entailment can be§approximated and modeled at the lexical level and§proposes a lexical reference subtask and a§correspondingly derived dataset. The book further§proposes a general probabilistic setting that casts§the applied notion of textual entailment in§probabilistic terms. This proposed setting may§provide a unifying framework for modeling uncertain§semantic inferences from texts. Finally, the book§presents a novel acquisition algorithm to identify§lexical entailment relations from a single corpus§focusing on the extraction of verb paraphrases.