Mode of access: World Wide Web.

System requirements: Adobe Acrobat Reader.

Part of: Synthesis digital library of engineering and computer science.

Series from website.

Includes bibliographical references (p. 171-197).

1. Textual entailment -- 1.1 Motivation and rationale -- 1.2 The recognizing textual entailment task -- 1.2.1 The scope of textual entailment -- 1.2.2 The role of background knowledge -- 1.2.3 Textual entailment versus linguistic notion of entailment -- 1.2.4 Extending entailment recognition with contradiction detection -- 1.2.5 The challenge and opportunity of RTE -- 1.3 Applications of textual entailment solutions -- 1.3.1 Question answering -- 1.3.2 Relation extraction -- 1.3.3 Text summarization -- 1.3.4 Additional applications -- 1.4 Textual entailment evaluation -- 1.4.1 RTE-1 through RTE-5 -- 1.4.2 RTE-6 and RTE-7 -- 1.4.3 Other evaluations of textual entailment technology -- 1.4.4 Future directions for entailment evaluation --

2. Architectures and approaches -- 2.1 An intuitive model for RTE -- 2.2 Levels of representation in RTE systems -- 2.2.1 Lexical-level RTE -- 2.2.2 Structured representations for RTE -- 2.3 Inference in RTE systems -- 2.3.1 Similarity-based approaches -- 2.3.2 Alignment-focused approaches -- 2.3.3 "Proof Theoretic" RTE -- 2.3.4 Hybrid approaches -- 2.4 A conceptual architecture for RTE systems -- 2.4.1 Preprocessing -- 2.4.2 Enrichment -- 2.4.3 Candidate alignment generation -- 2.4.4 Alignment selection -- 2.4.5 Classification -- 2.4.6 Main decision-making approaches -- 2.5 Emergent challenges -- 2.5.1 Knowledge acquisition bottleneck: acquiring rules -- 2.5.2 Noise-tolerant RTE architectures --

3. Alignment, classification, and learning -- 3.1 An abstract scheme for textual entailment decisions -- 3.2 Generating candidates and selecting alignments -- 3.2.1 Anchors: linking texts and hypotheses -- 3.2.2 Formalizing candidate alignment generation and alignment -- 3.3 Classifiers, feature spaces, and machine learning -- 3.4 Similarity feature spaces -- 3.4.1 Token-level similarity features -- 3.4.2 Structured similarity features -- 3.4.3 Entailment trigger feature spaces -- 3.4.4 Rewrite rule feature spaces -- 3.4.5 Discussion -- 3.5 Learning alignment functions -- 3.5.1 Learning alignment from gold-standard data -- 3.5.2 Learning entailment with a latent alignment --

4. Case studies -- 4.1 Edit distance-based RTE -- 4.1.1 Open source tree edit-based RTE system -- 4.1.2 Tree edit distance with expanded edit types -- 4.2 Logical representation and inference -- 4.2.1 Representation -- 4.2.2 Logical inference with abduction -- 4.2.3 Logical inference with shallow backoff system -- 4.3 Transformation-based approaches -- 4.3.1 Transformation-based approach with integer linear programming -- 4.3.2 Syntactic transformation with linguistically motivated rules -- 4.3.3 Syntactic transformation with a probabilistic calculus -- 4.3.4 Syntactic transformation with learned operation costs -- 4.3.5 Natural logic -- 4.4 Alignment-focused approaches -- 4.4.1 Learning alignment selection independently of entailment -- 4.4.2 Hand-coded alignment function -- 4.4.3 Leveraging multiple alignments for RTE -- 4.4.4 Aligning discourse commitments -- 4.4.5 Latent alignment inference for RTE -- 4.5 Paired similarity approaches -- 4.6 Ensemble systems -- 4.6.1 Weighted expert approach -- 4.6.2 Selective expert approach -- 4.7 Discussion --

5. Knowledge acquisition for textual entailment -- 5.1 Scope of target knowledge -- 5.2 Acquisition from manually constructed knowledge resources -- 5.2.1 Mining computation-oriented knowledge resources -- 5.2.2 Mining human-oriented knowledge resources -- 5.3 Corpus-based knowledge acquisition -- 5.3.1 Distributional similarity methods -- 5.3.2 Co-occurrence-based methods -- 5.3.3 Acquisition from parallel and comparable corpora -- 5.4 Integrating multiple sources of evidence -- 5.4.1 Integrating multiple information sources -- 5.4.2 Simultaneous global learning of multiple rules -- 5.5 Context sensitivity of entailment rules -- 5.6 Concluding remarks and future directions --

6. Research directions in RTE -- 6.1 Development of better/more flexible preprocessing tool chain -- 6.2 Knowledge acquisition and specification -- 6.3 Open source platform for textual entailment -- 6.4 Task elaboration and phenomenon-specific RTE resources -- 6.5 Learning and inference: efficient, scalable algorithms -- 6.6 Conclusion --

A. Entailment phenomena -- Bibliography -- Authors' biographies.

Abstract freely available; full-text restricted to subscribers or individual document purchasers.

Compendex

INSPEC

Google scholar

Google book search

In the last few years, a number of NLP researchers have developed and participated in the task of Recognizing Textual Entailment (RTE). This task encapsulates Natural Language Understanding capabilities within a very simple interface: recognizing when the meaning of a text snippet is contained in the meaning of a second piece of text. This simple abstraction of an exceedingly complex problem has broad appeal partly because it can be conceived also as a component in other NLP applications, from Machine Translation to Semantic Search to Information Extraction. It also avoids commitment to any specific meaning representation and reasoning framework, broadening its appeal within the research community. This level of abstraction also facilitates evaluation, a crucial component of any technological advancement program. This book explains the RTE task formulation adopted by the NLP research community, and gives a clear overview of research in this area. It draws out commonalities in this research, detailing the intuitions behind dominant approaches and their theoretical underpinnings. This book has been written with a wide audience in mind, but is intended to inform all readers about the state of the art in this fascinating field, to give a clear understanding of the principles underlying RTE research to date, and to highlight the short- and long-term research goals that will advance this technology.

Also available in print.

Title from PDF t.p. (viewed on August 14, 2013).