Search

We describe the development of an advanced retrieval engine to search information in libraries of chemical patents. We use the international TREC (Text Retrieval Conferences) evaluation framework to develop and assess an original search strategy combining a standard vector-space model, a network of co-citations between patents, and a strategy of standardization/expansion of chemical named-entities based on natural language processing. Our basic engine obtains an average accuracy of 0.067. The most significant precision gain is provided by the use of co-citations (+168%), while other contents, in particular ICP codes, do not improve retrieval effectiveness of the engine. The official TREC performance of our engine (ranked #1, with a mean average precision approaching 20%) emphasize the role of document contents as opposed to technological expertise in document retrieval. [ABSTRACT FROM AUTHOR]

This paper aims at understanding the reasons why variants of the derivatives of HAEMA — Greek for blood — occur in medical vocabulary. Aspects of variation are seen from a terminological standpoint, which may differ from that prevailing in lexicography or in natural language processing (NLP). We consider as variants of a term not only the forms that are spelt in a similar way, but also its synonyms, i.e. all the forms that have a similar definition. However, since hyponyms of a given term differ by their definitions, i.e. have a separate meaning, we treat them as separate terms: variation becomes part of term formation. The examples in English come from medical dictionaries, a glossary in haematology and MEDLINE. Several types of variation are analysed: graphic, morphological, morphosyntactic, syntactic and lexical, together with their possible meanings, and case studies of the use of some of the variants in context are presented. [ABSTRACT FROM AUTHOR]

This article explores various technologies involved in the dissemination of scholarly literature in multiple languages; it presents a range of technologies that are deemed necessary for effective, equitable multilingual information management, together with the current level of maturity of each technology. This work stems from the recognition of multilingualism in today's scholarly production worldwide, in contradiction with the observed unilingual, English predominance. The digital context is at the source of many transformations in the way information is disseminated and it offers new possibilities in this area. The presentation suggests many directions for future research, many of which pertain to tools for natural language processing; the latter are often non-existent for languages other than major European languages (due to many, non-technological reasons). The article also identifies a number of professional groups interested in this question, in order to illustrate the potential for growth as well as foreseeable repercussions in many spheres of scholarly activities. [ABSTRACT FROM AUTHOR]