Skip to main content
European Commission logo print header

Article Category

Content archived on 2022-11-17

Article available in the following languages:

EN

LRE results of call

A total of 88 proposals were registered in response to the October 1992 call for proposals published in the context of Linguistic Research and Engineering (LRE), Area 6 of the specific programme in the field of telematic systems in areas of general interest. The proposal was ...

A total of 88 proposals were registered in response to the October 1992 call for proposals published in the context of Linguistic Research and Engineering (LRE), Area 6 of the specific programme in the field of telematic systems in areas of general interest. The proposal was supplemented by an information package for supporting material available on request from the Commission. This included the technical background document, guidelines on how to make proposals, a general information document outlining the purpose and background to the Community LRE action, plus a model contract serving as a basis for shared-cost contracts. Some 2,100 information packages were supplied to a broad range of organizations based in 20 countries, including IT companies, SMEs, private research laboratories, universities, consultancy firms and government agencies. Requests for information were received directly or via the national delegations (the Telematics Management Committee) and other national focal points such as science and technology agencies, regional authorities and chambers of commerce. Workshops and information days giving advice on the call were held in Bonn, Paris, Pisa and Madrid. The response to the second call was very positive. Of the 88 proposals, six were rejected on formal grounds of eligibility, leaving 82 valid proposals for technical evaluation by experts. These 82 proposals involved a total of some 400 participants from 300 different organizations based in 17 countries. Distribution was balanced fairly evenly between universities (53%) and companies and RTD centres (45%). In total, the proposals requested a Community financial contribution of close to ECU 69 million: more than seven times the funds available. After assessment, 16 proposals were recommended for funding. Of these, it is likely that two will be merged during the negotiation phase. The accepted proposals represent a total of 2,000 man-months of work involving 90 project partners in all Community Member States and in several EFTA countries. Individual projects accepted within the three priority sub-areas addressed within the framework of LRE include: 1(a). Generic Research (Computational semantics): - FRACAS: To develop a common semantic framework within which it will be possible to integrate work from different traditions of formal semantics. 1(b). Generic Research (Assessment): - TSNLP: To define a methodology for the design of diagnostic test suites for NLP systems and to develop test suite fragments as well as an automatic test suite generation tool; - TEMAA: to define an assessment methodology for multilingual authoring aids, especially spelling and grammar checkers; - SQUALE: To exploit experience gained by the proposers in the US DARPA programme with a view to developing a multilingual evaluation paradigm for speech recognizer quality assessment. 2. Common Methods, tools and Language Resources. Collaborative Resources: - MULTAC To develop multilingual parallel corpora tagged according to TEI guidelines along with the associated tools for corpus annotation, manipulation and analysis, and prototypes for extraction of multi-word terms and construction of lexica for machine translation systems; - EUROTEXT: The objective of this project is to create an integrated set of text corpora and related tools and databases for human and machine use; - EUROCOCOSDA: To ensure a significant European contribution to the recently created world-wide group for speech databases and speech systems assessment (COCOSDA) and to ensure coordination at European level of several actions being undertaken within the group; - RELATOR: Aims to set up and operate a prototype network of repositories of written and spoken language data, rules and tools. - 3(a). Pilot and Demonstration Projects (Machine aided translation): - ANTHEM: To develop a prototype system capable of translating medical diagnosis and to encode them into an internationally recognized classification scheme. - 3(b). Pilot and Demonstration Projects (Advanced Office Automation Tools): - COMPASS: Aims to exploit machine readable versions of bilingual dictionaries to build the context sensitive look-up component of an interactive text comprehension aid for advanced language learners; - GIST: To combine knowledge representation and NLP techniques to develop a multilingual text generation system for instructional documents in the field of social security; - SECC: To develop a prototype grammar/style checker for controlled English for use in an industrial environment; - SIFT: Combined information and NPL techniques to build a demonstrator that will select passages in SGML texts (instructions in computer manuals) in response to a query formed in natural language; - CRISTAL: Aims to develop multilingual document retrieval facilities through a natural language interface to news articles in French on the basis of a concept-based dictionary; - RENOS; Intends to improve the performance of text retrieval techniques by building a concept network for the legal sub-language (fiscal and general legislation in several EC languages). The tools developed are to exploit both statistical and rule-based approaches; - TRANSTERM: A toolbox for the creation, normalization and customization of terminological resources and their integration into application specific lexicons. The tools will be validated through field tests carried out by industrial partners. Compared the first call for proposals, which addressed a wide range of themes, the second call has taken a more selective and integrated approach to technology development and application building, with the main emphasis on a number of focused, complementary actions addressing a few key issues. Half of the projects liable to receive EC funds are concerned with the building of practical applications based on existing knowledge. It is, however, worth noting that RTD tasks are present in all projects, together with integration and validation of innovative technologies. Despite the limited budget available, the growing interest and participation of industry, including SMEs, shows the success of LRE in contributing to an increased awareness across the Community of the economic impact of language modelling and engineering activities and in stimulating the emergence of an European language infrastructure. No further calls for proposals will be launched in the framework of LRE. The preparation of Language Engineering actions within the Fourth Framework Programme is now underway.

Related articles