Representing Meaning with a Combination of Logical and Distributional Models

Dados Bibliográficos

ID12156352
AutoresIslam Beltagy (The University of Texas at Austin, autor correspondente), Stephen Roller (The University of Texas at Austin), Pengxiang Cheng (0000-0001-5997-705X, The University of Texas at Austin), Katrin Erk (0000-0002-2888-6961, The University of Texas at Austin), Raymond J Mooney (0000-0002-4504-0490, The University of Texas at Austin)
Ano2016
Volume42
Fascículo4
Páginas763-808
Data de publicação2016-09-28
Peer ReviewedSim
Open AccessSim
TipoARTICLE
PeriódicoComputational Linguistics (JOURNAL)
Identificadores do periódicoISSN: 0891-2017 • E-ISSN: 1530-9312
EditoraAssociation for Computational Linguistics (PUBLISHER • US)
DOI10.1162/coli_a_00266
OpenAlexW2963869850
IdiomaEN
Citações recebidas4
Referências citadas47

NLP tasks differ in the semantic information they require, and at this time no single semantic representation fulfills all requirements. Logic-based representations characterize sentence structure, but do not capture the graded aspect of meaning. Distributional models give graded similarity ratings for words and phrases, but do not capture sentence structure in the same detail as logic-based approaches. It has therefore been argued that the two are complementary. We adopt a hybrid approach that combines logical and distributional semantics using probabilistic logic, specifically Markov Logic Networks. In this article, we focus on the three components of a practical system: 1 1) Logical representation focuses on representing the input problems in probabilistic logic; 2) knowledge base construction creates weighted inference rules by integrating distributional information with other sources; and 3) probabilistic inference involves solving the resulting MLN inference problems efficiently. To evaluate our approach, we use the task of textual entailment, which can utilize the strengths of both logic-based and distributional representations. In particular we focus on the SICK data set, where we achieve state-of-the-art results. We also release a lexical entailment data set of 10,213 rules extracted from the SICK data set, which is a valuable resource for evaluating lexical entailment systems. 2

Focus (optics · Inference · Knowledge base · Logical consequence · Natural language processing · Probabilistic logic · Programming language · Representation (politics · Semantics (computer science · Sentence · Set (abstract data type · Textual entailment · Advanced Text Analysis Techniques · Computer Science · Natural Language Processing Techniques · Topic Modeling · Artificial Intelligence