Francesco Ronzano wins the first Language Technologies Hackathon at the 4YFN event
Francesco Ronzano wins the first Language Technologies Hackathon at the 4YFN event
The researcher of the Natural Language Processing research group developed the application TextDigester, which generates summaries automatically.
TextDigester, an application that generates summaries automatically, was the winner of the first Language Technologies Hackathon, held on 27 February in Barcelona, within the framework of the event 4 Years From Now linked to the Mobile World Congress.
The development was created by Francesco Ronzano, a researcher with the Natural Language Processing research group (TALN), of the Department of Information and Communication Technologies at UPF.
The application allows analysing textual content of HTML pages such as a digital newspaper, XML documents such as RSS, and JSON objects such as the Tweets on Twitter. Having detected the language, it runs an analysis of the lexis and semantics of the documents and uses this information to implement several methods to generate summaries in different languages.
TextDigester allows its users to choose from several summary approaches that can be easily created and customized. The application is implemented as a stand-alone Java library that integrates natural language processing frameworks like Freeling, GATE, Deeplearning4j and SUMMA. The code of the application is available publicly on GitHub.
The winners of the competition were chosen in two separate votes with the same weighting, one featuring the participants in the Hackathon, and the other, the jury. The two agreed on the three top finishers.
The other finalists developed applications covering a variety of issues, including chatbots, tools to support learning, or information retrieval. Participants in the competition included research centres, like Vicomtech, and technological companies such as IBM Everis.
The jury was comprised of María Fernández (Red.es), Horacio Rodríguez and Luis Padró (Polytechnic University of Catalonia), Núria Bel (Pompeu Fabra University), and David Pérez (SESIAD)
This first Hackathon is part of Agenda Digital’s Plan Nacional de Tecnologías del Lenguaje - national plan for the advancement of language technology- and aims to promote applications related to natural language processing and automatic translation.