Consortium as a whole. The MOLTO Consortium has four partners, of which three are academic and one industrial. The consortium was built with a great care to match the vision of MOLTO and provide the competences needed without too much overlap. The result of the process is a consortium that also covers a representative set of five different countries, diverging both geographically as in terms of language families: ▇▇▇▇▇-Ugric, Germanic, Romance, and Slavic. An essential question in a multilingual project like MOLTO is to find a sufficient basis of developers and testers for the different languages. Here, the consortium itself comes a long way towards the goal: its key persons alone have proficiency in at least ten languages. More languages are available in the immediate vicinity: in the Department of Computer Science and Engineering of the coordinating site UGOT alone, 30 nationalities are represented. The main competences and responsibilities of each partner can be summarized as follows: UGOT, University of Gothenburg, Coordinator. UGOT has a leading competence in multilingual grammar formalisms and grammar resources, and the group coordinates the collaborative open- source development of GF. In MOLTO, UGOT is responsible for the design and implementation of grammar development tools (WP2) and the availability of linguistic resources. UGOT also provides technical help in integrating GF with the translation tools (WP3), the Knowledge Engineering (WP4), and statistical methods (WP5). Moreover, UGOT has the leading role in the Cultural Heritage case study, where it builds on its previous competence on the domain, as well as collaboration with Gothenburg City Museum. In WP7, UGOT will develop the grammars needed in the hybrid model. As Coordinator of MOLTO, UGOT has the main responsibility for management (WP1) and dissemination (WP10). UHEL, University of Helsinki. ▇▇▇▇ has competence in human translator training and translation tools, as well as in grammar development and ontologies. UHEL is therefore the main responsible partner for translator’s tools (WP3) and requirements and evaluation (WP9). The group has both research and practical experience with CAT and MT tools (taught Trados tools since 1995), including involvement in national R&D projects where CAT tools have been or are developed (MLIS Lingmachine, Masterin TM/MT system, Multilingual Workbench). The group was also involved in the development of mathematical GF grammars for the WebALT project. UPC, Universitat Politecnica de Catalunya. From UPC, two groups are involved: applied mathematics, the main party responsible in WP6, and computational linguistics, the main party responsible in WP5 and WP7. In WP5, UPC will provide the SMT technology needed for the research in this package, coordinate the corpora compilation/alignment, and develop the grammar/statistical-based combined MT models Wide experience in the construction and evaluation of Statistical Machine Translation systems, machine learning of statistical natural language parsers, and a combination of different sources of linguistic information in the construction of SMT systems. In WP6, UPC is the main responsible partner, developing grammars for natural language generation and parsing, collecting exercise samples and validation, and implementing automated mathematical reasoning. In WP7, UPC will test the hybrid model developed in WP5 on the patent corpus provided by EPO. The group has ample teaching experience in mathematics at university level in several curricula and was the main designer of mathematical GF grammars for the WebALT project. ICREA (Institucio Catalana de Recerca i Estudis Avancats Fundacio Privada) is a research organization functioning as a third party under UPC. Ontotext, Ontotext AD. Ontotext will make the spectrum of its semantic technology and competence available to the MOLTO consortium by leading WP4 (Knowledge Engineering). They will deliver research and development of two-way grammar - ontology interoperability, infrastructure for knowledge modeling, semantic indexing and retrieval and ontology modeling and alignment of structured data sources. Ontotext will contribute to the retrieval, navigation and visualization of knowledge and ontology-grammar interoperability in the grammar development IDE (WP2) and the use case systems (WP6, WP7, WP8), developing the prototypes for two of the use cases (museums and patents). The company will heavily participate in the dissemination and exploitation (WP10) activities, on the forums for semantic technology it usually sponsors and maintain the MOLTO Web portal with live demos. The retrieval and MT outcomes of MOLTO will be integrated in the products of the company. Mxw, Matrixware GmbH.41 An information services company, Mxw is the leader of the patent MT and retrieval case study in WP7. Mxw specializes in analysis and retrieval of technical content in the intellectual property domain (patents, prior-art). It invests heavily in research and development in the areas of information retrieval, information extraction, large data set visualization, and recently semantic annotation and search. It is the aggregator and developer of the Alexandria patent repository for patents and patent metadata. Alexandria provides an extensible global storage facility for high-quality scientific, technical and business information, which includes a substantial collection of international patents (at the time of writing this the patent collection numbers over 70 million documents). Mxw will define the use case requirements (WP9), provide parallel corpora (WP5) and participate in the evaluation (WP9) and feasibility studies (WP10). Mxw will disseminate through their academic and industrial partnerships and appropriate events. Here is a list of the recent previous experiences of the key persons of MOLTO in European projects: • UGOT: TYPES (FP6-2002-IST-C), CLARIN (FP7-RI-2122230), TALK (IST-507802), DHomme (IST-2000-26280), TRINDI (LE4-8314) • UHEL: WebALT (eContent 22253) • UPC: ACQUILEX (Esprit BRA 3030), ACQUILEX–II (Esprit BRA 7315), EuroWordNet (LE-4003), NAMIC (IST-1999-12392), MEANING (IST-2001-34460), FAME (IST-2001- 28323), LC-STAR (IST-2001-32216), CHIL (IST 506909), HOPS (IST? 507967), WebALT (eContent 22253). • Ontotext: TAO (IST-2004-026460), TripCom (IST-4-027324-STP), RASCALLI (IST- 27596-2004), SEKT (IST-2003-506826) • Mxw: ePatent, eMage The extension adds two more countries, The Netherlands and Switzerland, and adds Dutch, German, French, and Estonian as new languages covered by the Consortium staff. The main competences and responsibilities of each partner can be summarized as follows: UGOT, acting as coordinator, will develop the GF application grammars needed for work 41Mxw left the Consortium in April 2010. The tasks of Mxw have been reallocated among the remaining partners. package WP11, and assist in preparing the infrastructure in WP12. UZH and BI will respectively carry out most of the work planned in work-packages WP11 and WP12 since they are leading experts in Attempto, CNL, Semantic Wiki and in industrial workflows and user perspective. UHEL will participate in building the semantic wiki system in WP11. ▇▇▇▇ will also be responsible for evaluating and assessing the MOLTO technologies applied to the new work- packages in WP9(E). Key persons of MOLTO-Enlarged EU have experience in European projects related to the area of MOLTO. Most relevant are: UZH: REWERSE (Reasoning on the Web with Rules and Semantics, IST-506779 FP6) BI:▇▇▇▇▇▇ (Multilingual ontologies for networked knowledge, FP7 248458-STREP) ongoing.
Appears in 1 contract
Sources: Grant Agreement
Consortium as a whole. The MOLTO Consortium has four five partners, of which three are academic and one two industrial. The consortium was built with a great care to match the vision of MOLTO and provide the competences needed without too much overlap. The result of the process is a consortium that also covers a representative set of five different countries, diverging both geographically as in terms of language families: ▇▇▇▇▇-Ugric, Germanic, Romance, and Slavic. An essential question in a multilingual project like MOLTO is to find a sufficient basis of developers and testers for the different languages. Here, the consortium itself comes a long way towards the goal: its key persons alone have proficiency in at least ten languages. More languages are available in the immediate vicinity: in the Department of Computer Science and Engineering of the coordinating site UGOT alone, 30 nationalities are represented. The main competences and responsibilities of each partner can be summarized as follows: UGOT, University of Gothenburg, Coordinator. UGOT has a leading competence in multilingual grammar formalisms and grammar resources, and the group coordinates the collaborative open- open-source development of GF. In MOLTO, UGOT is responsible for the design and implementation of grammar development tools (WP2) and the availability of linguistic resources. UGOT also provides technical help in integrating GF with the translation tools (WP3), the Knowledge Engineering (WP4), and statistical methods (WP5). Moreover, UGOT has the leading role in the Cultural Heritage case study, where it builds on its previous competence on the domain, as well as collaboration with Gothenburg City Museum. In WP7, UGOT will develop the grammars needed in the hybrid model. As Coordinator of MOLTO, UGOT has the main responsibility for management (WP1) and dissemination (WP10). UHEL, University of Helsinki. ▇▇▇▇ has competence in human translator training and translation tools, as well as in grammar development and ontologies. UHEL is therefore the main responsible partner for translator’s tools (WP3) and requirements and evaluation (WP9). The group has both research and practical experience with CAT and MT tools (taught Trados tools since 1995), including involvement in national R&D projects where CAT tools have been or are developed (MLIS Lingmachine, Masterin TM/MT system, Multilingual Workbench). The group was also involved in the development of mathematical GF grammars for the WebALT project. UPC, Universitat Politecnica de Catalunya. From UPC, two groups are involved: applied mathematics, the main party responsible in WP6, and computational linguistics, the main party responsible in WP5 and WP7WP5. In WP5, UPC will provide the SMT technology needed for the research in this package, coordinate the corpora compilation/alignment, and develop the grammar/statistical-based combined MT models Wide experience in the construction and evaluation of Statistical Machine Translation systems, machine learning of statistical natural language parsers, and a combination of different sources of linguistic information in the construction of SMT systems. In WP6WP 6, UPC is the main responsible partner, developing grammars for natural language generation and parsing, collecting exercise samples and validation, and implementing automated mathematical reasoning. In WP7, UPC will test the hybrid model developed in WP5 on the patent corpus provided by EPO. The group has ample teaching experience in mathematics at university level in several curricula and was the main designer of mathematical GF grammars for the WebALT project. ICREA (Institucio Catalana de Recerca i Estudis Avancats Fundacio Privada) is a research organization functioning as a third party under UPC. Ontotext, Ontotext AD. Ontotext will make the spectrum of its semantic technology and competence available to the MOLTO consortium by leading WP4 (Knowledge Engineering). They will deliver research and development of two-two- way grammar - ontology interoperability, infrastructure for knowledge modeling, semantic indexing and retrieval and ontology modeling and alignment of structured data sources. Ontotext will contribute to the retrieval, navigation and visualization of knowledge and ontology-grammar interoperability in the grammar development IDE (WP2) and the use case systems (WP6, WP7, WP8)-8), developing the prototypes for two of the use cases (museums and patents). The company will heavily participate in the dissemination and exploitation (WP10) activities, on the forums for semantic technology it usually sponsors and maintain the MOLTO Web portal with live demos. The retrieval and MT outcomes of MOLTO will be integrated in the products of the company. Mxw, Matrixware GmbH.41 An information services company, Mxw is the leader of the patent MT and retrieval case study in WP7. Mxw specializes in analysis and retrieval of technical content in the intellectual property domain (patents, prior-art). It invests heavily in research and development in the areas of information retrieval, information extraction, large data set visualization, and recently semantic annotation and search. It is the aggregator and developer of the Alexandria patent repository for patents and patent metadata. Alexandria provides an extensible global storage facility for high-quality scientific, technical and business information, which includes a substantial collection of international patents (at the time of writing this the patent collection numbers over 70 million documents). Mxw will define the use case requirements (WP9), provide parallel corpora (WP5) and participate in the evaluation (WP9) and feasibility studies (WP10). Mxw will disseminate through their academic and industrial partnerships and appropriate events. Here is a list of the recent previous experiences of the key persons of MOLTO in European projects: • UGOT: TYPES (FP6-2002-IST-C), CLARIN (FP7-RI-2122230), TALK (IST-507802), DHomme (IST-2000-26280), TRINDI (LE4-8314) • UHEL: WebALT (eContent 22253) • UPC: ACQUILEX (Esprit BRA 3030), ACQUILEX–II (Esprit BRA 7315), EuroWordNet (LE-4003), NAMIC (IST-1999-12392), MEANING (IST-2001-34460), FAME (IST-2001- 28323), LC-STAR (IST-2001-32216), CHIL (IST 506909), HOPS (IST? 507967), WebALT (eContent 22253). • Ontotext: TAO (IST-2004-026460), TripCom (IST-4-027324-STP), RASCALLI (IST- 27596-2004), SEKT (IST-2003-506826) • Mxw: ePatent, eMage The extension adds two more countries, The Netherlands and Switzerland, and adds Dutch, German, French, and Estonian as new languages covered by the Consortium staff. The main competences and responsibilities of each partner can be summarized as follows: UGOT, acting as coordinator, will develop the GF application grammars needed for work 41Mxw left the Consortium in April 2010. The tasks of Mxw have been reallocated among the remaining partners. package WP11, and assist in preparing the infrastructure in WP12. UZH and BI will respectively carry out most of the work planned in work-packages WP11 and WP12 since they are leading experts in Attempto, CNL, Semantic Wiki and in industrial workflows and user perspective. UHEL will participate in building the semantic wiki system in WP11. ▇▇▇▇ will also be responsible for evaluating and assessing the MOLTO technologies applied to the new work- packages in WP9(E). Key persons of MOLTO-Enlarged EU have experience in European projects related to the area of MOLTO. Most relevant are: UZH: REWERSE (Reasoning on the Web with Rules and Semantics, IST-506779 FP6) BI:▇▇▇▇▇▇ (Multilingual ontologies for networked knowledge, FP7 248458-STREP) ongoing.
Appears in 1 contract
Sources: Grant Agreement