Weighted Abduction for Discourse Processing Based on Integer Linear Programming

Naoya Inoue; Ekaterina Ovchinnikova; Kentaro Inui; Jerry Hobbs

doi:10.1016/B978-0-12-398532-3.00002-6

Weighted Abduction for Discourse Processing Based on Integer Linear Programming

Naoya Inoue, Ekaterina Ovchinnikova, Kentaro Inui, Jerry Hobbs

INF - System Information Sciences

Research output: Chapter in Book/Report/Conference proceeding › Chapter

4 Citations (Scopus)

Abstract

This chapter explores the logical framework called weighted abduction as applied to solving discourse-processing tasks. Weighted abduction incorporates a cost propagation mechanism allowing us to estimate the likelihood of the obtained abductive proofs. We use a tractable implementation of weighted abduction based on Integer Linear Programming and a large knowledge base generated automatically. We first perform an experiment on plan recognition using the dataset originally developed for Ng and Mooney's system [39]. Then we apply our discourse processing pipeline for predicting whether one text fragment logically entails another one (Recognizing Textual Entailment task). The study we describe is the first attempt to apply tractable inference-based natural language processing on a large scale.

Original language	English
Title of host publication	Plan, Activity, and Intent Recognition
Subtitle of host publication	Theory and Practice
Publisher	Elsevier Inc.
Pages	33-55
Number of pages	23
ISBN (Print)	9780123985323
DOIs	https://doi.org/10.1016/B978-0-12-398532-3.00002-6
Publication status	Published - 2014 Mar

Keywords

Integer Linear Programming
Natural language understanding
Planning
Recognizing Textual Entailment
Weighted abduction

ASJC Scopus subject areas

Computer Science(all)

Access to Document

10.1016/B978-0-12-398532-3.00002-6

Fingerprint Dive into the research topics of 'Weighted Abduction for Discourse Processing Based on Integer Linear Programming'. Together they form a unique fingerprint.

Cite this

Inoue, N., Ovchinnikova, E., Inui, K., & Hobbs, J. (2014). Weighted Abduction for Discourse Processing Based on Integer Linear Programming. In Plan, Activity, and Intent Recognition: Theory and Practice (pp. 33-55). Elsevier Inc.. https://doi.org/10.1016/B978-0-12-398532-3.00002-6

@inbook{6d060d4c4993445daf6b8f25274a8814,

title = "Weighted Abduction for Discourse Processing Based on Integer Linear Programming",

abstract = "This chapter explores the logical framework called weighted abduction as applied to solving discourse-processing tasks. Weighted abduction incorporates a cost propagation mechanism allowing us to estimate the likelihood of the obtained abductive proofs. We use a tractable implementation of weighted abduction based on Integer Linear Programming and a large knowledge base generated automatically. We first perform an experiment on plan recognition using the dataset originally developed for Ng and Mooney's system [39]. Then we apply our discourse processing pipeline for predicting whether one text fragment logically entails another one (Recognizing Textual Entailment task). The study we describe is the first attempt to apply tractable inference-based natural language processing on a large scale.",

keywords = "Integer Linear Programming, Natural language understanding, Planning, Recognizing Textual Entailment, Weighted abduction",

author = "Naoya Inoue and Ekaterina Ovchinnikova and Kentaro Inui and Jerry Hobbs",

year = "2014",

month = mar,

doi = "10.1016/B978-0-12-398532-3.00002-6",

language = "English",

isbn = "9780123985323",

pages = "33--55",

booktitle = "Plan, Activity, and Intent Recognition",

publisher = "Elsevier Inc.",

}

TY - CHAP

T1 - Weighted Abduction for Discourse Processing Based on Integer Linear Programming

AU - Inoue, Naoya

AU - Ovchinnikova, Ekaterina

AU - Inui, Kentaro

AU - Hobbs, Jerry

PY - 2014/3

Y1 - 2014/3

N2 - This chapter explores the logical framework called weighted abduction as applied to solving discourse-processing tasks. Weighted abduction incorporates a cost propagation mechanism allowing us to estimate the likelihood of the obtained abductive proofs. We use a tractable implementation of weighted abduction based on Integer Linear Programming and a large knowledge base generated automatically. We first perform an experiment on plan recognition using the dataset originally developed for Ng and Mooney's system [39]. Then we apply our discourse processing pipeline for predicting whether one text fragment logically entails another one (Recognizing Textual Entailment task). The study we describe is the first attempt to apply tractable inference-based natural language processing on a large scale.

AB - This chapter explores the logical framework called weighted abduction as applied to solving discourse-processing tasks. Weighted abduction incorporates a cost propagation mechanism allowing us to estimate the likelihood of the obtained abductive proofs. We use a tractable implementation of weighted abduction based on Integer Linear Programming and a large knowledge base generated automatically. We first perform an experiment on plan recognition using the dataset originally developed for Ng and Mooney's system [39]. Then we apply our discourse processing pipeline for predicting whether one text fragment logically entails another one (Recognizing Textual Entailment task). The study we describe is the first attempt to apply tractable inference-based natural language processing on a large scale.

KW - Integer Linear Programming

KW - Natural language understanding

KW - Planning

KW - Recognizing Textual Entailment

KW - Weighted abduction

UR - http://www.scopus.com/inward/record.url?scp=84902906097&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902906097&partnerID=8YFLogxK

U2 - 10.1016/B978-0-12-398532-3.00002-6

DO - 10.1016/B978-0-12-398532-3.00002-6

M3 - Chapter

AN - SCOPUS:84902906097

SN - 9780123985323

SP - 33

EP - 55

BT - Plan, Activity, and Intent Recognition

PB - Elsevier Inc.

ER -