The tutorial will also include a practical session in which we will see how to disambiguate and link real news text as well as bags of words of various kinds in different languages. We will also show how Babelfy can work in a language-agnostic setting, in which the text language does not need to be recognized by the system. Babelfy is based on random walks with restart to weight the importance of related concepts and a densest subgraph heuristic which determines the most interconnected interpretation graph for a given input text. This tutorial will introduce Babelfy, a unified knowledge-based approach that leverages BabelNet to jointly perform word sense disambiguation and entity linking in arbitrary languages, with performance on both tasks on a par with, or surpassing, those of task-specific state-of-the-art supervised systems. Multilingual Joint Word Sense Disambiguation and Entity Linking in any language with Babelfy The tutorial will also include a practical session in which we will see how to query BabelNet 3.0 in Java and via SPARQL. With its 13.8 million synsets and 2 billion RDF triples, BabelNet is a core component of the Linked Open Data cloud and a powerful engine for virtually any Natural Language Processing task in desperate need of wide-coverage lexical semantics in arbitrary languages. BabelNet 3.0 also integrates the Wikipedia Bitaxonomy, therefore providing a full taxonomization of both concepts and named entities. This tutorial will introduce BabelNet 3.0, the largest multilingual encyclopedic dictionary and semantic network, which covers 271 languages and provides both lexicographic and encyclopedic knowledge for all the open-class parts of speech, thanks to the seamless integration of WordNet, Wikipedia, Wiktionary, OmegaWiki, Wikidata and the Open Multilingual WordNet. BabelNet 3.0: a core for Linguistic Linked Data and NLP
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |