Project description DEENESFRITPL Training machine learning models for natural language tasks The internet contains vast amounts of data and information, written and audiovisual, and in many different languages. There is a growing need to take advantage of this largely untapped resource. The EU-funded SELMA project will address the ingestion and monitoring of large amounts of data. The project will systematically train machine learning models for natural language tasks and use these models to monitor data streams, aiming to improve multilingual media monitoring and news content production. The project will ultimately advance the state of the art in language modelling, machine translation and speech recognition and synthesis. Show the project objective Hide the project objective Objective SELMA builds a continuous deep learning multilingual media platform using extreme analytics. Large amounts of multilingual text and speech data are available in the internet, but the potential to fully take advantage of this data has remained largely untapped. Recent advances in deep learning and transfer learning have opened the door to new possibilities – in particular integrating knowledge from these large unannotated datasets into plugable models for tackling machine learning tasks. The aim of the Stream Learning for Multilingual Knowledge Transfer (SELMA) is to address three tasks: ingest large amounts of data and continuously train machine learning models for several natural language tasks; monitor these data streams using such models to improve multilingual Media Monitoring (use case 1); and improve the task of multilingual News Content Production (use case 2), thereby closing the loop between content monitoring and production. SELMA has eight goals: 1. Enable processing of massive video and text data streams in a distributed and scalable fashion 2. Develop new methods for training unsupervised deep learning language models in 30 languages 3. Enable knowledge transfer across tasks and languages, supporting low-resourced languages 4. Develop novel data analytics methods and visualizations to facilitate the media monitoring decision-making process 5. Develop an open-source platform to optimize multilingual content production in 30 languages 6. Fine-tune deep learning models from user feedback, reducing recurring errors 7. Ensure a sustainable exploitation of the SELMA platform 8. Encourage active user involvement in the platform. Achieving these aims requires advancing the state of the art in multiple technologies (transfer learning, language modelling, speech recognition, machine translation, summarization, speech synthesis, named entity linking, learning from user feedback), while building upon previous project results and existing services. Fields of science natural sciencescomputer and information sciencesinternetnatural sciencescomputer and information sciencesartificial intelligencemachine learningtransfer learningnatural sciencescomputer and information sciencesartificial intelligencemachine learningdeep learning Keywords stream learning learning from user feedback extreme analytics summarization multilingualism natural language understanding speech to text machine translation named entity recognition Programme(s) H2020-EU.2.1.1. - INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies - Information and Communication Technologies (ICT) Main Programme Topic(s) ICT-51-2020 - Big Data technologies and extreme-scale analytics Call for proposal H2020-ICT-2018-20 See other projects for this call Sub call H2020-ICT-2020-1 Funding Scheme RIA - Research and Innovation action Coordinator DEUTSCHE WELLE Net EU contribution € 821 812,50 Address KURT SCHUMACHER STRASSE 3 53113 Bonn Germany See on map Region Nordrhein-Westfalen Köln Bonn, Kreisfreie Stadt Activity type Public bodies (excluding Research Organisations and Secondary or Higher Education Establishments) Links Contact the organisation Opens in new window Participation in EU R&I programmes Opens in new window HORIZON collaboration network Opens in new window Total cost € 821 812,50 Participants (4) Sort alphabetically Sort by Net EU contribution Expand all Collapse all AVIGNON UNIVERSITE France Net EU contribution € 599 281,25 Address RUE LOUIS PASTEUR 74 84029 Avignon Cedex 01 See on map Region Provence-Alpes-Côte d’Azur Provence-Alpes-Côte d’Azur Vaucluse Activity type Higher or Secondary Education Establishments Links Contact the organisation Opens in new window Participation in EU R&I programmes Opens in new window HORIZON collaboration network Opens in new window Total cost € 599 281,25 LATVIJAS UNIVERSITATES MATEMATIKAS UN INFORMATIKAS INSTITUTS Latvia Net EU contribution € 576 250,00 Address RAINA BULVARIS 29 1459 Riga See on map Region Latvija Latvija Rīga Activity type Research Organisations Links Contact the organisation Opens in new window Website Opens in new window Participation in EU R&I programmes Opens in new window HORIZON collaboration network Opens in new window Total cost € 576 250,00 PRIBERAM INFORMATICA SA Portugal Net EU contribution € 749 412,50 Address ALAMEDA D AFONSO HENRIQUES 41 2 1000-123 Lisboa See on map SME The organization defined itself as SME (small and medium-sized enterprise) at the time the Grant Agreement was signed. Yes Region Continente Área Metropolitana de Lisboa Área Metropolitana de Lisboa Activity type Private for-profit entities (excluding Higher or Secondary Education Establishments) Links Contact the organisation Opens in new window Participation in EU R&I programmes Opens in new window HORIZON collaboration network Opens in new window Total cost € 749 412,50 FRAUNHOFER GESELLSCHAFT ZUR FORDERUNG DER ANGEWANDTEN FORSCHUNG EV Germany Net EU contribution € 705 750,00 Address HANSASTRASSE 27C 80686 Munchen See on map Region Bayern Oberbayern München, Kreisfreie Stadt Activity type Research Organisations Links Contact the organisation Opens in new window Website Opens in new window Participation in EU R&I programmes Opens in new window HORIZON collaboration network Opens in new window Total cost € 705 750,00
