Open-Minded Coreference Resolution Sieve Based on Answer Set Programming

OmSieve project

Project Description

In the OmSieve project we will apply methods of Answer Set Programming to coreference resolution. We want to deal with linguistic ambiguities on all levels of representation in a way that is more open-minded than existing methods.

Answer Set Programming (ASP) is a general purpose logic programming formalism that supports comfortable representation of knowledge, nonmonotonic reasoning processes, and reasoning with hybrid knowledge bases. In an ASP logic program we describe (i) a set of potential solutions, (ii) relationships between concepts in the solution, and (iii) constraints on solutions. Given such a representation an ASP solver (a software tool) computes those solutions that adhere to the specified relationships and constraints. An example is that a robot wants to find the way out of a labyrinth: potential solutions are sequences of movements from the starting point, the location of the robot is a concept which depends on its movements, and constraints are that the robot cannot walk through walls and that it must arrive outside of the labyrinth. ASP solvers can find all solutions to such problems and they are engineered to find these solutions efficiently. Moreover ASP supports hybrid reasoning which means that some relationships between concepts can be described outside the ASP logic, in external reasoning modules, for example we can obtain the labyrinth's wall locations from a website, and the ASP solver will request only those data that are necessary to find a solution from the website.

Coreference resolution is the Computer Linguistics task of finding out which phrases of a natural language discourse refer to the same entity in the world. For example in the sentence “He said to the people: 'I need your help'” the task is to find out that “he” and “I” refers to the same entity (the speaker), furthermore “the people” and “your” refers to the same entity (the listeners). Coreference resolution is challenging: noun phrases can refer to the same entity for various reasons, they can be synonyms, hypernyms, or hyponyms, or they can be coreferent because of background knowledge and discourse information (e.g., “my brother”, “John”, and “the king” can be coreferent due to contextual information). Coreference of pronouns with noun phrases is additionally challenging because linguistic constraints (e.g., agreement of gender or number features) are not sufficient for pointing out a single correct entity for each pronoun — again background knowledge can be necessary to identify the correct solution.

The Sieve technique is a recently introduced and successful coreference resolution method. In this project we want to use Answer Set Programming to create an open-minded Sieve (OmSieve) where Sieve modules can reconsider internal decisions and interact with other modules in a bidirectional manner.

Project Team:

Project Outcomes:

J  Journal Article  C  Conference Paper  E  Editorship  O  Other  

Publications 2018:

O . Marmara Turkish Coreference Corpus and Coreference Resolution Baseline. Technical Report, Marmara University & TU Wien, , Version 2. [  ]

J . Answer Set Programming applied to Coreference Resolution and Semantic Similarity. KI - Künstliche Intelligenz 32 (2), pages 207-208, , DOI: 10.1007/s13218-018-0539-7. [  ]

J . Adjudication of Coreference Annotations via Answer Set Optimization. Journal of Experimental & Theoretical Artificial Intelligence 30 (4), pages 525-546, , DOI: 10.1080/0952813X.2018.1456793. [  ]

Publications 2017:

C . Adjudication of Coreference Annotations via Answer Set Optimization. In: Logic Programming and Nonmonotonic Reasoning (LPNMR), volume 10377 of Lecture Notes in Computer Science, pages 343-357, , Best Application Paper, DOI: 10.1007/978-3-319-61660-5_31.

O . Coreference Resolution Sieve based on Answer Set Programming. Masters Thesis, Marmara University, . [  ]

Publications 2016:

O , , , and . Turkish Coreference Annotation Manual (V2). . [  ]

C . Adjudication of Coreference Annotations via Finding Optimal Repairs of Equivalence Relations. In: International Workshop on Experimental Evaluation of Algorithms for Solving Problems with Combinatorial Explosion (RCRA), volume 1745 of CEUR Workshop Proceedings, pages 57-71, . [  ]

J , , , , , and . A Model Building Framework for Answer Set Programming with External Computations. Theory and Practice of Logic Programming 16 (04), pages 418-464, , arXiv:1507.01451 [cs.AI], DOI: 10.1017/S1471068415000113. [  ] [  ]

J , , and . A systematic analysis of levels of integration between high-level task planning and low-level feasibility checks. AI Communications 29 (2), pages 319-349, , DOI: 10.3233/AIC-150697. [  ] [  ]

C , , and . Problem Solving Using the HEX Family. In: Computational Models of Rationality - Essays dedicated to Gabriele Kern-Isberner on the occasion of her 60th birthday, pages 150-174, . [  ] [  ]

Publications 2015:

O , , and . Problem Solving Using the HEX Family. Technical Report INFSYS RR-1843-15-07, Institut für Informationssysteme, TU Wien, , Favoritenstraße 9-11, A-1040 Vienna. [  ]

C  and . Answer Set Application Programming: a Case Study on Tetris. In: International Conference on Logic Programming (ICLP), Technical Communications, volume 1433 of CEUR Workshop Proceedings, . [  ] [  ]

O  and . Turkish Coreference Annotation Manual (V1). . [  ]

Impressum: Medieninhaber Peter Schüller (Privatperson), Fasangartengasse 57 1130 Wien AUSTRIA. Email: Telefon: +43(0)69910963525