Fundamentals in natural language processing: a proposal for extraction bigrams

• Autor(a) principal: Silva, Edson Marchetti da
• Data de Publicação: 2014
• Outros Autores: Souza, Renato Rocha
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Encontros Bibli
• Texto Completo: https://periodicos.ufsc.br/index.php/eb/article/view/1518-2924.2014v19n40p1
• topic: Extração de expressões multipalavras; Medidas de associação estatísticas; Heudet; Multiword expression extraction; Measures of association statistics; Heudet

It is common sense that the written text is an important way of to register information and currently much of this information content is available in digital form. However, in general, the computers consider a text is a string that have not significance. The area of Natural Language Processing (PLN) has been engaged in extracting meaning from text. Accordingly this paper presents a review of this issue and proposes an automated method that uses a deterministic heuristic called Heudet which aims extract bigram of the text. The goal is to extract the meaning of the text identifing a set of multiword expressions (MWE). The results were better compared to those using up the techniques of statistical association measures obtained from the software ngram Statistics Package (NSP).