Term Extractor Research Articles

Funcionalidades del sistema de gestión de corpus, GECO, para la elaboración de diccionarios de especialidad

The use of computational tools for lexicographical work in the creation of specialty dictionaries is undeniable. The process starts from the design and constitution of the specialty corpus, through the extraction of terms, to the automated extraction of definitions, which is a work of computational lexicography. This paper sketches in first place the elements to be considered in specialty corpus constitution. Subsequently the GECO system is described, a manager to upload the corpus, as well as the system applications for dictionaries construction such as the automatic term extractor, the concordances viewer and the definitional contexts extractor. This approach shows an effective solution to the problems in the process of choosing terms and extracting definitions from a large corpus, which significantly reduces the processing time of the data obtained in order to improve the creation of a specialized dictionary.

Bilingual term recognition revisited

The paper describes LUIZ, a bilingual term recognition system that has been developed for the Slovene-English language pair. The system is a hybrid term extractor using morphosyntactic patterns and statistical ranking to propose domain-specific expressions for each of the two languages, whereupon translation equivalents between the languages are identified using the innovative bag-of-equivalents approach. This simple but effective method is based on the Twente word aligner to obtain a lexicon of single word translation pairs and their probability scores, which is then used to identify correspondences between multi-word terms. The bilingual term recognition system has been tested and evaluated on three parallel subcorpora from the tourism, accounting and military domain. Average precision of the term alignment component is 0.83, whereby only fully equivalent and domain-relevant terms were counted as positives. Another advantage of the described approach is the fact that we successfully detect term variants and multiple translations of a candidate multi-word term. Since our term alignment method does not require sentence-aligned corpora it can be used with comparable corpora, provided we already have a domain-specific lexicon or dictionary of single-word correspondences. The paper concludes with some thoughts on the users of term recognition systems and their needs based on our observations from the online version of the system.

Automating the compilation of specialized dictionaries

In this paper, we study how single-word term extraction and bilingual lexical alignment can be used and combined to assist terminologists when they compile bilingual specialized dictionaries. Two specific tools — namely a term extractor called TermoStat and a sentence and lexical aligner called Alinea — are tested in a specific project the aim of which is the development of an English-French dictionary on climate change. We analyze the results of lexical alignment based on a typology of terminology equivalents. We first extracted French candidate terms that were then submitted to the lexical aligner. The results show that the use of these tools proves to be a valuable asset for compiling bilingual dictionaries. Most equivalents provided by the aligner were valid and the tool was able to locate several valid English equivalents (some of which were structurally different) for candidate terms.

Evaluation of terms and term extraction systems

Term extraction may be defined as a text mining activity whose main purpose is to obtain all the terms included in a text of a given domain. Since the eighties, and mainly due to the rapid scientific advances as well as the evolution of the communication systems, there has been a growing interest in obtaining the terms found in written documents. A number of techniques and strategies have been proposed for satisfying this requirement. At present it seems that term extraction has reached a maturity stage. Nevertheless, many of the systems proposed fail to qualitatively present their results, almost every system evaluates its abilities in an ad hoc manner (if any, many times). Often, the authors do not explain their evaluation methodology; therefore comparisons between different implementations are difficult to draw. In this paper, we review the state-of-the-art of term extraction systems evaluation in the framework of natural language systems evaluation. The main approaches are presented, with a focus on their limitations. As an instantiation of some ideas for overcoming these limitations, the evaluation framework is applied to YATE, a hybrid term extractor.

Automatic extraction of candidate nomenclature terms using the doublet method

BackgroundNew terminology continuously enters the biomedical literature. How can curators identify new terms that can be added to existing nomenclatures? The most direct method, and one that has served well, involves reading the current literature. The scholarly curator adds new terms as they are encountered. Present-day scholars are severely challenged by the enormous volume of biomedical literature. Curators of medical nomenclatures need computational assistance if they hope to keep their terminologies current. The purpose of this paper is to describe a method of rapidly extracting new, candidate terms from huge volumes of biomedical text. The resulting lists of terms can be quickly reviewed by curators and added to nomenclatures, if appropriate. The candidate term extractor uses a variation of the previously described doublet coding method. The algorithm, which operates on virtually any nomenclature, derives from the observation that most terms within a knowledge domain are composed entirely of word combinations found in other terms from the same knowledge domain. Terms can be expressed as sequences of overlapping word doublets that have more specific meaning than the individual words that compose the term. The algorithm parses through text, finding contiguous sequences of word doublets that are known to occur somewhere in the reference nomenclature. When a sequence of matching word doublets is encountered, it is compared with whole terms already included in the nomenclature. If the doublet sequence is not already in the nomenclature, it is extracted as a candidate new term. Candidate new terms can be reviewed by a curator to determine if they should be added to the nomenclature. An implementation of the algorithm is demonstrated, using a corpus of published abstracts obtained through the National Library of Medicine's PubMed query service and using "The developmental lineage classification and taxonomy of neoplasms" as a reference nomenclature.ResultsA 31+ Megabyte corpus of pathology journal abstracts was parsed using the doublet extraction method. This corpus consisted of 4,289 records, each containing an abstract title. The total number of words included in the abstract titles was 50,547. New candidate terms for the nomenclature were automatically extracted from the titles of abstracts in the corpus. Total execution time on a desktop computer with CPU speed of 2.79 GHz was 2 seconds. The resulting output consisted of 313 new candidate terms, each consisting of concatenated doublets found in the reference nomenclature. Human review of the 313 candidate terms yielded a list of 285 terms approved by a curator. A final automatic extraction of duplicate terms yielded a final list of 222 new terms (71% of the original 313 extracted candidate terms) that could be added to the reference nomenclature.ConclusionThe doublet method for automatically extracting candidate nomenclature terms can be used to quickly find new terms from vast amounts of text. The method can be immediately adapted for virtually any text and any nomenclature. An implementation of the algorithm, in the Perl programming language, is provided with this article.

A rough set-based case-based reasoner for text categorization

This paper presents a novel rough set-based case-based reasoner for use in text categorization (TC). The reasoner has four main components: feature term extractor, document representor, case selector, and case retriever. It operates by first reducing the number of feature terms in the documents using the rough set technique. Then, the number of documents is reduced using a new document selection approach based on the case-based reasoning (CBR) concepts of coverage and reachability. As a result, both the number of feature terms and documents are reduced with only minimal loss of information. Finally, this smaller set of documents with fewer feature terms is used in TC. The proposed rough set-based case-based reasoner was tested on the Reuters21578 text datasets. The experimental results demonstrate its effectiveness and efficiency as it significantly reduced feature terms and documents, important for improving the efficiency of TC, while preserving and even improving classification accuracy.