The MWN.PT WordNet for Portuguese: Projection, Validation, Cross-lingual Alignment and Distribution

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 4859–4866 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 4859 The MWN.PT WordNet for Portuguese: Projection, Validation, Cross-lingual Alignment and Distribution Ant´onio Branco, Sara Grilo, M´arcia Bolrinha, Chakaveh Saedi, Ruben Branco, Jo˜ao Silva, Andreia Querido, Rita de Carvalho, Rosa del Gaudio, Mariana Avel˜as, Clara Pinto University of Lisbon NLX—Natural Language and Speech Group, Department of Informatics Faculdade de Ciˆencias Campo Grande, 1749-016 Lisboa, Portugal antonio.branco@di.fc.ul.pt Abstract The objective of the present paper is twofold, to present the MWN.PT WordNet and to report on its construction and on the lessons learned with it. The MWN.PT WordNet for Portuguese includes 41,000 concepts, expressed by 38,000 lexical units. Its synsets were manually validated and are linked to semantically equivalent synsets of the Princeton WordNet of English, and thus transitively to the many wordnets for other languages that are also linked to this English wordnet. To the best of our knowledge, it is the largest high quality, manually validated and cross-lingually integrated, wordnet of Portuguese distributed for reuse. Its construction was initiated more than one decade ago and its description is published for the first time in the present paper. It follows a three step <projection, validation with alignment, completion> methodology consisting on the manual validation and expansion of the outcome of an automatic projection procedure of synsets and their hypernym relations, followed by another automatic procedure that transferred the relations of remaining semantic types across wordnets of different languages. Keywords: Wordnet, ontology, Portuguese, lexical semantics, lexical semantic network 1. Introduction Lexical semantic networks are pervasive in Natural Lan- guage Processing (NLP) as they play a key role in virtu- ally all major applications, e.g. in information retrieval, in- formation extraction, machine translation, summarization, and question answering, among others. In what concerns NLP specific tasks, ontologies have also been of paramount importance, for instance, in word sense disambiguation, anaphora resolution, and in semantic role assignment, to name just a few. A great deal of research has been devoted to devising meth- ods that support the development of ontologies in a way as much automatic as possible. Such methods tend to be of quite disparate nature from a technological point of view, and their chances of success rely very much on the quality of the linguistic resources from which the ontologies are to be extracted. These resources may range from highly struc- tured sources of linguistic information (e.g. pre-existing ontologies or bilingual dictionaries) to totally unprocessed materials (e.g. collections of raw texts). The construction of ontologies of significant size requires a considerable amount of human effort, and the methods aiming at supporting their (semi-)automatic extraction or development are usually said to be addressing the so-called knowledge acquisition bottleneck. However, regardless of how successful these methods may be in tackling this issue, it is worth noting that they will be finding another major bottleneck later on, which could be termed the knowledge validation bottleneck, so much so that …

Content truncated to 500 words