Towards a Seamless Integration of Word Senses into Downstream NLP Applications

Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 1857–1869 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1170 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 1857–1869 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1170 Towards a Seamless Integration of Word Senses into Downstream NLP Applications Mohammad Taher Pilehvar2, Jose Camacho-Collados1, Roberto Navigli1 and Nigel Collier2 1Department of Computer Science, Sapienza University of Rome 2Department of Theoretical and Applied Linguistics, University of Cambridge 1{collados,navigli}@di.uniroma1.it 2{mp792,nhc30}@cam.ac.uk Abstract Lexical ambiguity can impede NLP sys- tems from accurate understanding of se- mantics. Despite its potential benefits, the integration of sense-level information into NLP systems has remained understudied. By incorporating a novel disambiguation algorithm into a state-of-the-art classifica- tion model, we create a pipeline to inte- grate sense-level information into down- stream NLP applications. We show that a simple disambiguation of the input text can lead to consistent performance im- provement on multiple topic categoriza- tion and polarity detection datasets, par- ticularly when the fine granularity of the underlying sense inventory is reduced and the document is sufficiently large. Our re- sults also point to the need for sense rep- resentation research to focus more on in vivo evaluations which target the perfor- mance in downstream NLP applications rather than artificial benchmarks. 1 Introduction As a general trend, most current Natural Language Processing (NLP) systems function at the word level, i.e. individual words constitute the most fine-grained meaning-bearing elements of their in- put. The word level functionality can affect the performance of these systems in two ways: (1) it can hamper their efficiency in handling words that are not encountered frequently during train- ing, such as multiwords, inflections and deriva- tions, and (2) it can restrict their semantic under- standing to the level of words, with all their am- biguities, and thereby prevent accurate capture of the intended meanings. The first issue has recently been alleviated by techniques that aim to boost the generalisation power of NLP systems by resorting to sub-word or character-level information (Ballesteros et al., 2015; Kim et al., 2016). The second limitation, however, has not yet been studied sufficiently. A reasonable way to handle word ambiguity, and hence to tackle the second issue, is to semantify the input text: transform it from its surface-level semantics to the deeper level of word senses, i.e. their intended meanings. We take a step in this di- rection by designing a pipeline that enables seam- less integration of word senses into downstream NLP applications, while benefiting from knowl- edge extracted from semantic networks. To this end, we propose a quick graph-based Word Sense Disambiguation (WSD) algorithm which allows high confidence disambiguation of words without much computation overload on the system. We evaluate the pipeline in two downstream NLP ap- plications: polarity detection and topic categoriza- tion. Specifically, we use a classification model based on Convolutional Neural Networks which has been shown to be …

Content truncated to 500 words