Distributional analysis applied to terminology extraction

Abstract

This paper presents the first results of a new method for terminology extraction based on distributional analysis. The intuition behind the algorithm is that single or multi-word lexical units that refer to specialised concepts will show a characteristic co-occurrence pattern, described as a tendency to appear in the same contexts with other conceptually related terms. E.g. the termfluoxetinewill systematically appear in the same sentences with other related terms such asdepression, serotonin reuptake inhibitor, obsessive–compulsive disorderand others. Of course, terms will co-occur with general vocabulary units as well, but not with a characteristic pattern as when a conceptual relation holds. Experimental evaluation of this method was conducted in a corpus of psychiatry journals from Spain and Latin America, and concluded that the results are significantly better than other methods.

Descripción de rasgos léxicos de textos profesionales producidos por estudiantes de derecho y medicina

“…Por su parte, para el análisis de la terminología se empleó la herramienta Termout (Nazar, 2016), que nos permitió reconocer los términos prototípicos de cada una de las disciplinas. En tanto, las siglas y abreviaturas se reconocieron a partir de un análisis manual, que nos permitió identificar su frecuencia en los textos que componen cada subcorpus.…”

“…Para analizarlo, hemos trabajado con variadas técnicas y herramientas que nos han permitido explorar las características de los textos del corpus a nivel léxico. Las dimensiones exploradas y las respectivas herramientas utilizadas en cada caso son las siguientes: terminología, analizada con Termout (Nazar, 2016); colocaciones, tiempos verbales y marcadores de subjetividad, explorados con SINLP (Crossley et al, 2014) y la Plataforma de Análisis Cuantitativo de Textos Español (PACTE); y, por último, siglas y abreviaturas, identificadas a partir de un análisis manual.…”

Application-based Evaluation of Automatic Terminology Extraction

“…In terms of ATE, designation of a word or a phrase as a term is not a simple binary decision. ATE result is presented as a continuum, in the form of a list of candidates ranked according to the score [11].…”

Machine Learning in Terminology Extraction from Czech and English Texts

“…There have been dozens of studies of automatic term recognition (or automatic term extraction, computer-assisted term acquisition etc., Yang 1986; Kageura, Umino 1996;Heid 1998Heid /1999Gamper, Stock 1998/1999Lossio--Ventura et al 2014;Nazar 2016) since the late 1980s. The cooperation of linguists, terminologists and computer scientists brings more and more successful and accurate methods for identifying terms in written texts.…”