QTLeap: a highly successful project !
QTLeap project was highly successful . Its achievements received a highly positive assessment by external reviewers. Check here for a condensed presentation of its key results.
QTLeap project was highly successful . Its achievements received a highly positive assessment by external reviewers. Check here for a condensed presentation of its key results.
The Deep Machine Translation Workshop (DMTW), organized by the QTLeap Project, took place at the University of Lisbon, on October 21, 2016.
The DMTW goal was to bring together researchers and students working on machine translation approaches and technology using “deep understanding”.
QTLeap Advisory Board Meeting, Rectory of the University of Lisbon
On October 20th 2016, the QTLeap Project organized the last Advisory Board of Potential Users (ABPU) meeting.
It took place at the University of Lisbon and brought together the members of the Advisory Board and the consortium.
The meeting was very productive. Its focus was on the project final results and evaluation as well as on the follow-up cooperation actions.
September 26th, 2016, Museu do Oriente, Lisbon, Portugal
António Branco, professor and researcher at the University of Lisbon in the area of science and language technology (Department of Informatics, Faculty of Sciences), is one of the invited speakers to the discussion panel “Plural Portuguese”, at the Conference “30 Years of Portuguese Language in the EU“.
This conference is co-organized by the European Commission, the European Parliament and the Camões – Instituto da Cooperação e da Língua, I.P to highlight the presence of the Portuguese language in the European institutions in the last three decades and takes place in Museu do Oriente, in Lisbon, September 26, 2016.
The discussion panel is led by Nicolau Santos, Director of the newspaper “Expresso”, and also has the participation of Fernando Frutuoso de Melo, Head of the Civil House of the President of the Republic and former Director-General at the European Commission, Ana Paula Laborinho, President of the Camões – Instituto da Cooperação e da Língua I.P., José Ribeiro e Castro, Portuguese lawyer and politician with a connection with the media field and the Portuguese language, and Célio Conceição, professor and researcher at the University of Algarve.
On 10 September 2016 the QTLeap team organized the Workshop on Deep Language Processing for Quality Machine Translation (DeepLP4QMT).
Invited talks were given by: Francis Bond (Nanyang Technological University) and Josef van Genabith (DFKI).
The workshop enjoyed big interest also from non-project members.
On 10 September 2016, the QTLeap team organizes a Workshop on Deep Language Processing for Quality Machine Translation (DeepLP4QMT) in conjunction with AIMSA conference, Varna, Bulgaria.
At the First Conference on Machine Translation Workshop, held on 11-12 August 2016 in conjunction with ACL, Berlin, the QTLeap team participated with a shared-task paper “SMT and Hybrid systems of the QTLeap project in the WMT16 IT-task”.
The poster, which was received with interest by the community, was presented by Martin Popel (CUNI).
PROPOR – International Conference on the Computational Processing of Portuguese is the main scientific meeting in the area of language and speech technologies for the Portuguese language and on the basic and applied research issues related to this language.
The 12th edition of the PROPOR Conferences took place in Tomar, Portugal, 13-15 July 2016.
PROPOR2016 was organized and supported by the QTLeap project.
Co-organized by QTLeap project
Follow up of the ELRC Lisbon Workshop
It was a very successful workshop highly participated with presentations on language resources suitable to support statistical machine translation from/into Portuguese and on processing tools for their preparation.
The workshop also contributed to exchange ideas and disseminate best practices that help to foster the ELRC and CEF.AT initiatives.
The organization of META-FORUM 2016 was supported by the University of Lisbon, under the responsibility of António Branco, and by the QTLeap project.
The META-FORUM 2016 in Lisbon was an event highly participated, full of interesting presentations, discussions, keynotes, invited speeches and posters.
António Branco, Coordinator of QTLeap Project, participated in the Opening and Closing Sessions of META-FORUM 2016 Conference and also presented the key results and achievements of QTLeap.
QTLeap Project took part on the first Poster Session of META-FORUM 2016.
Photos by Sergio Miguel (NC Produções)
META-FORUM 2016 takes place in Lisbon, Portugal, on 4-5 July, 2016.
It is an international conference on powerful and innovative language technologies for the multilingual information society, the data value chain and the European multilingual Digital Single Market.
The organization of META-FORUM 2016 is supported by the University of Lisbon, under the responsibility of António Branco, and by the QTLeap project.
The closing address is delivered by Secretary of State for Science, Technology and Higher Education of the Portuguese Government, Maria Fernanda Rollo.
META-FORUM is a conference series organised by META-NET, a European network of excellence that consists of 60 research centres in 34 European countries.
Key results from the QTLeap project were presented in the LREC conference (May 25, 2016) and were granted the outstanding opportunity of being the first presentation of the first session in the conference.
In the picture Aljoscha Burchardt is taking care of the presentation.
The 10th edition of LREC, the International Conference on Language Resources and Evaluation, will take place in Portorož (Slovenia), 23-28 May 2016.
LREC is the largest conference on tools and resources for Natural Language Processing.
QTLeap will be present at LREC 2016 with dozens of presentations by several members of the QTLeap team.
The presentations will address issues related to the QTLeap Project such as Machine Translation and Lexical Semantics.
The Executive Secretariat of the Community of Portuguese Language Countries (CPLP) will celebrate on May 5, 2016: “The Day of the Portuguese Language and CPLP Culture“.
The opening ceremony under the theme “CPLP 20 years – Cultural Diversity That Unites Us” will be chaired by the Executive Secretary of CPLP, Ambassador Murade Murargy and will take place in Lisbon, in Conde of Penafiel Palace.
António Branco, Coordinator of QTLeap Project, will participate in the opening ceremony, in a round table under the theme “Our Language, What Future?”. This round table moderated by Marisa Mendonça, Professor and Executive Director of the International Institute of the Portuguese Language (IILP), will also be attended by the Writer João Melo and the Professor Nataniel Ngomane of the Portuguese Language Bibliographic Fund.
QTLeap Advisory Board Meeting
Last April 29, 2016, the QTLeap Project organized another meeting of the Advisory Board of Potential Users (ABPU).
The 3rd Advisory Board Meeting took place at the DFKI Project Office, in Berlin and brought together all the members of Advisory Board and the consortium.
It was a very successful meeting with an increasing number of usage cases that are uptaking the results of the project.
QTLeap team: General Project Meeting, Berlin
On April 28, 2016 the general project meeting of the consortium took place at DFKI Project Office, in Berlin.
The partners reported their achievements, discussed future plans and project ideas and decided next steps in the project work plan.
The 3rd Advisory Board of Potential Users (ABPU) will take place at DFKI Project Office, in Berlin.
This Strategic Project Meeting, which brings together all members of the Advisory Board of Potential Users (ABPU), and the QTLeap consortium, aims to observe the progress behind the project to exploit its business potential.
Follow up of the ELRC Lisbon Workshop
The workshop seeks to contribute to improve on this state of affairs by helping to map both available parallel datasets suitable to feed statistical machine translation systems and available language processing tools useful for their preparation.
The workshop invites submissions on resources and tools for any language that fit into the stated aim of this workshop.
Workshop call for papers information available on: Corpora and Tools for Processing Corpora.
As part of the celebration of the 725 years of its foundation, the University of Coimbra organized the International Congress “Portuguese Language: a language for the future” held from December 2 – 4, 2015.
On the second day of the Congress, António Branco participated as a speaker at the first plenary table entitled “The Portuguese language in the digital age and in the information society“.
10 minute interview (in Portuguese) with António Branco broadcasted by the national radio RTP – Antena2 on the programme “Páginas da Língua Portuguesa”, by Miguel Roque Dias, in November 29, 2015.
To listen to the full programme visit the programme site (after skipping ads, after minute 19:38).
The 2nd year General Project Meeting of QTLeap Project took place at the University of Lisbon, Faculty of Sciences, on November 26-27, 2015. The partners discussed their achievements on Pilot 2, which introduced lexical semantics into the MT systems. Also, the activities for year 3 were planned.
QTLeap partners presented their contributions at WMT15 Poster Session.
“DFKI’s experimental hybrid MT system”, by Eleftherios Avramidis, Maja Popovic and Aljoscha Burchardt.
“New Language Pairs in TectoMT”, by Ondřej Dušek, Luís Gomes, Michal Novák, Martin Popel and Rudolf Rosa.
The WMT 2015 – Workshop on Statistical Machine Translation was held on September 17-18, in Lisbon, Portugal, at the EMNLP 2015 – Conference on Empirical Methods in Natural Language Processing.
Abstracts: Since creating the first instantiation of the LLOD cloud diagram in 2012, LLOD has attracted a lot of activity, we have reached an agreement on vocabularies for many aspects of language resources and the number of resources included is continuously on the rise. This growth is documented, for example, by declaring LLOD “the new hot topic in our (= language resource) community” (Nicoletta Calzolari, LREC-2014 closing session). But with substantial amounts of data being available, the focus of activity in the LLOD community is slowly shifting from resource creation to applications of Linguistic Linked Open Data. The primary promise of providing open, but heterogeneously structured and scattered language resources in a more interoperable way has been fulfilled, and it facilitates using and re-using existing language resources in novel contexts. Beyond this, innovative *LLOD-based* applications for common problems in Natural Language Processing, Digital Humanities and linguistics are on the horizon.
The QTLeap project is orgnizing the Deep Machine Translation Workshop, who is held in September 3-4, 2015.
This is the first workshop on “Deep Machine Translation”. Its aim is to bring together researchers and students working on machine translation approaches and technology using “deep understanding”.
The S2MT-Workshop on Semantics-Driven Statistical Machine Translation: Theory and Practice was held on July 30, in Beijing, collocated with ACL-IJCNLP’2015 conference, the top conference in the area of language technology.
António Branco delivered a keynote speech with the title “Machine Translation and Deep Language Engineering Approaches“.
He was also one of the discussants in the panel that closed the workshop, with the title “Semantics and Statistical Machine Translation: Gaps and Challenges”, together with Chris Quirk (Microsoft), Eduard Hovy (Carnegie Mellon University), Percy Liang (Stanford University), and Quoc V. Le (Google).
From 29 June to 1 July 2015 a Developers Workshop and a Project Meeting were organized in San Sebastian, Basque Country. Both events were hosted by UPV/EHU.
Each partner group presented the progress of their work and ongoing collaborations related to WP5 “Lexical semantics: Linking and resolving”.
The participants made important decisions about the strategies and specific steps to follow for the successful completion of Pilot 2 in September 2015.
From 29 June to 1 July 2015 the partner UPV/EHU is hosting a Developers Workshop and a Project Meeting in San Sebastian, Basque Country.
Both events are dedicated to Pilot 2, which is set under the motto: “getting it lexically linked and resolved”.
The topics of discussion include the usage of Word Sense Disambiguation, Named Entity Detection and Disambiguation as well as Linked Open Data for better Machine Translation.
The first Deep Machine Translation Workshop will take place in Prague, Czech Republic, on 3rd-4th September 2015.
It aims at bringing together researchers and students working on machine translation approaches and technology using “deep understanding”.
Submissions are invited on aspects of deep language analysis, generation and natural language understanding in direct or indirect relation with Machine Translation.
- YRRSDS (Workshop on Spoken Dialogue Systems for PhDs, PostDocs & New Researchers), August 31 – September 1, 2015.
- SIGdial (Meeting on Discourse and Dialogue), September 2-4, 2015.
- MTM (Machine Translation Marathon), September 7-12, 2015.
The winners of the prize QTLeap Best Paper Award – Maryam Aminian, Mahmoud Ghoneim, and Mona Diab, from The George Washington University, for “Unsupervised False Friend Disambiguation Using Contextual Word Clusters and Parallel Word Alignments” – were announced at the workshop.
Nora Aranberri delivering the Amazon voucher of €500 to Maryam Aminian, for the winning of the Best Paper Award sponsored by the QTLeap project.
The SSST-9 – Ninth Workshop on Syntax, Semantics and Structure in Statistical Translation collocated with NAACL 2015, took place in Denver, Colorado, June 4, 2015.
Nora Aranberri and Iñaki Alegria from the University of the Basque Country presented their contributions at SSST-9 workshop co-organized by QTLeap.
Josu Goikoetxea presented his poster “Random Walks and Neural Network Language Models on Knowledge Bases”, work done in collaboration with Eneko Agirre and Aitor Soroa.
At the workshop *SEM 2015, June 4-5 Eneko Agirre presented a task organized in SemEval-2015 Task 2: Semantic Textual Similarity, English, Spanish and Pilot on Interpretability, and a poster for the work “Combining Mention Context and Hyperlinks from Wikipedia for Named Entity Disambiguation” in collaboration with Ander Barrena and Aitor Soroa.
This year SSST-9 will award a best paper award among papers which advance MT using lexical semantics and deep language processing.
This award is sponsored by the QTLeap project. The winner of the prize will be announced at the workshop, and will win an Amazon voucher of €500.
The SSST-9 will take place June 4, 2015, Denver, Colorado, collocated with NAACL2015 – Conference of the North American Chapter of the Association for Computational Linguistics – Human Language Technologies.
Co-organized by QTLeap project and partially funded by the European Commission (FP7-ICT-2013-10-610516).
QTLeap project is present at the Riga Summit on the Multilingual Digital Single Market, which includes the META-FORUM conference, and is held in Riga, Latvia, April 27-29, 2015.
QTLeap project partners Hans Uszkoreit (DFKI, Germany), António Branco (University of Lisbon, Portugal) and Jan Hajič (Charles University in Prague, Czech Republic) present their contributions at META-FORUM conference that takes place as part of the Riga Summit on April 27. The summary of the programme is available at http://www.meta-forum.eu.
QTLeap Project at the Poster Session
On day two, on the Summit Plenary, Hans Uszkoreit, makes a presentation on the “Strategic Agenda for the Multilingual DSM: A Research Perspective”.
On the last day of the Riga Summit, it will be held the CEF conference “Towards a Connected Multilingual Europe”.
António Branco gives a presentation under the theme: “Synergy of National Activities in Creation of Language Resources for Multilingual Europe”.
2nd Advisory Board Meeting, Prague, Thursday, April 16, 2015
The 2nd Advisory Board of Potential Users (ABPU) took place at Faculty of Mathematics and Physics, in Charles University in Prague.
This meeting was attended by all the advisory board members, the invited external scientific experts and all the project members.
QTLeap project partners, FCUL and IICT-BAS, present at PARSEME 4th General Meeting in Valletta, Malta, 19-20 March 2015.
FCUL poster presentation: MWEs in the QTLeap Corpus of Online Helpdesk Interactions.
IICT-BAS poster presentation: Formalizing MultiWords as Catenae in a Treebank and in a Lexicon.
In April the QTLeap team is organizing a number of important events in Prague, Czech Republic.
All the events hosted by the Czech partner – CUNI take place at Faculty of Mathematics and Physics, Charles University.
On 14 and 15 April, a boost workshop on Pilot 2 is envisaged. This event is important with respect to the synchronization of the partners’ efforts towards the tasks in Pilot 2.
On 16 April, the second Advisory Board meeting will be held. All the members of the Advisory Board will participate in order to observe the progress behind the project and its growing application potential.
On 17 April, the general project meeting of the consortium is taking place. All the partners will report on the achievements and will discuss the current and next steps in the project work plan.
On 18 April, the Project Board meeting is scheduled.
Basque newspapers “El Diario Vasco”, “Noticias de Gipuzkoa”, “Noticias de Navarra” and “Deia” present first year results of QTLeap Project
Online newspaper “Bilbao Buenas Noticias BBN”, “Aprendemas”, “#languageandtranslation” and “Catalunya Vanguardista” report first year results of QTLeap Project
Machine translation and the QTLeap Project presented in “Euskadi Irratia” and “Onda Vasca” radio interview (2015/02/23)
Kepa Sarasola, researcher in QTLeap project from the University of the Basque Country (UPV/EHU) was interviewed for 10 minutes in the program Hiri Gorrian (listen from 1:12:00 to 1:22:45) in Euskadi Irratia, the major national radio station of the Basque Country. (2015/02/27, 17:26-17:37)
Kepa Sarasola, researcher in QTLeap project from the University of the Basque Country (UPV/EHU) was interviewed for 5 minutes in the program Euskadi Hoy Magazine in Onda Vasca, a major national radio station of the Basque Country. (2015/02/23, 13:32-13:37)
The 11th International Conference on Natural Language Processing (ICON-2014) took place at Goa University, Goa, India, during December, 18-21, 2014.
December, 19, 2014, Eleftherios Avramidis (DFKI) presented a publication of a joint work with Maja Popovic (DFKI), at the 11th International Conference on Natural Language Processing at Goa University in India.
Machine translation, in general, and QTLeap, in particular, had the full page of its “Science” section in the last edition of Expresso, January 10th, including figures of our extrinsic evaluation of Pilot 0, in an article with the title “The Dream of a Perfect Translation” (in Portuguese). Read full article here.
The Portuguese weekly newspaper Expresso is the top reference among the reputable Portuguese media, and one with the widest readership in Portugal, in its paper and online versions.
Expresso, January 10, 2015, “science” section, p. 35
The newspaper Expresso, the reference weekly newspaper in Portugal, interviewed António Branco and stressed the importance of QTLeap Project in an article published Friday, January 2, 2015.
The first Annual Meeting of the European Project QTLeap took place at Faculty of Science, University of Lisbon, on December 5-6, 2014.
The SSST-9 - ”Ninth Workshop on Syntax, Semantics and Structure in Statistical Translation” organized by Eneko AGIRRE and Nora ARANBERRI, University of the Basque Country together with Dekai Wu, Hong Kong University of Science and Technology (HKUST) and Marine Carpuat, National Research Council (NRC) Canada, takes place in Denver, Colorado, USA, in Jun, 4, 2015. The SSST-9 will award a best paper sponsored by the QTLeap Project.
Rudolf Rosa (CUNI) presents the poster “Depfix, a tool for automatic rule-based post-editing of SMT” at Machine Translation Marathon, September 8th-13th, 2014 in Trento.
The 10th Languages & The Media Conference, under the theme Smart Technologies, Smart Translations, took place in Berlin from November 5 – 7, 2014.
November 6, 2014, Aljoscha Burchardt (DFKI) gave a presentation entitled “An Analytic Approach to Machine Translation Evaluation and Improvement“.
Photo Credit: ICWE GmbH / Mark Bollhorst
RDA Europe Workshop about Data Sharing and Interoperability, October, 17, 2014, Brussels
The Coordinator of the QTLeap Project, António Branco participated in the RDA Europe Workshop about Data Sharing and Interoperability organized by RDA (Research Data Alliance).
European Day of Languages, September 26th, 2014, Museu do Oriente, Lisbon, Portugal
Within the scope of the celebration of the European Day of Languages, the Portuguese Parliament, the Department of Portuguese of the Directorate-General for Translation of the European Commission, the European Commission Representation in Lisbon, the Prosecutor General’s Office and the Camões – Institute of Cooperation and Language co-organize a conference entitled “Language: Translating the Future.”
António Branco participates in this Conference as Moderator of Panel 4: Preparation of the Portuguese Language in the Digital Age.
António Branco, project coordinator, was interviewed in the program “Days of the Future”, dedicated to science and technology, in Antena 1, a major national radio station of Portugal.
After skipping the initial seconds of advertising, the interview can be listened here, from minute 12 to minute 35 (in Portuguese).
The colleagues from the University of Lisbon are the local organizers of the 10th DELPH-IN meeting – Deep Linguistic Processing with HPSG. It will be held in Tomar from 14 July to 18 July 2014. More than 30 participants will share their experiences on the development of Deep Grammars for various languages. More info can be found at: http://moin.delph-in.net/TomarTop
During the LREC 2014 conference in Reykjavik, the QTLeap Project presented its new t-shirt model at the HLT Village.
QTLeap researchers are among the organizers of 9th SaLTMiL Workshop on “Free/open-Source Language Resources for the Machine Translation of Less-Resourced Languages”, co-located with LREC-2014, Reykjavik, Iceland, 27th May 2014.
QTLeap is present in the EACL2014 conference, in Gothenburg, April 26-30, 2014.
In the photos, participants consulting the QTleap leaflet during coffee break.
António Branco and Petya Osenova will present the main ideas behind the QTLeap project: QTLeap – Quality Translation with Deep Language Engineering Approaches as a poster at EAMT2014 – The Seventeenth Annual Conference of the European Association for Machine Translation, Dubrovnik, 16-18 June, 2014.
Ondřej Bojar (CUNI) co-organizes WMT2014 – 9th Workshop on Statistical Machine Translation to be held in Baltimore, USA, on 26-27 June 2014, collocated with ACL2014 – The Annual Conference of the Association for Computational Linguistics.
This year’s workshop will feature five shared tasks: a translation task, a quality estimation task, a task to test automatic evaluation metrics, a medical text translation task.
In addition to the shared tasks, the workshop will also feature scientific papers on topics related to MT.
Our project has a dissemination booth in the HLT Project Village. By attractive multimedia shows, flyers and direct conversations at the booth, QTLeap project members will present project goals and current progress.
António Branco, Kiril Simov and Petya Osenova are the QTLeap co-organizers of the LREC Workshop LDL2014 – 3rd Workshop on Linked Data in Linguistics: Multilingual Knowledge Resources and Natural Language Processing, 27 May 2014.
Kepa Sarasola is the QTLeap co-organizer of the LREC Workshop on Free/open-source language resources for the machine translation of less-resourced languages. UPV/EHU has a paper on Wikipedia and Machine Translation at the workshop. The Proceedings are available at: http://ixa2.si.ehu.es/saltmil/
CUNI, IICT-BAS and FCUL have papers at the main conference. Their focus is respectively on ‘stanfordizing’ 30 dependency treebanks’, improved constituency and dependency parsing models for Bulgarian, and HLT research agenda.
The talk reports on recent and ongoing work dedicated to analytical methods for a systematic combination of observed strengths of translation technologies. The focus will be on different ways of exploiting existing data on MT output and performance measures for system combination and for gaining insights on strengths and weaknesses of existing technologies.
Third Workshop on Hybrid Approaches to Translation (HyTra)
Machine Translation (MT) is a highly interdisciplinary and multidisciplinary field since it is approached from the point of view of human translators, engineers, computer scientists, mathematicians and linguists. This workshop aims at motivating the cooperation and interaction between them, and to foster innovative combinations between the two main MT paradigms: statistical and rule-based.
The Coordinator of the QTLeap Project, António Branco participated in the networking workshop of the European Project MLi -Towards a Multilingual Data & Services Infrastructure: Brainstorming on the future of Meta-Share and Needs in the Field. This collaborative meeting has been organized by ELDA/ELRA in Paris on April 7, 2014.
The Project Board with the External Experts meeting was organized by FCUL, the coordinator of the QTLeap Project and it took place in Lisbon, March 15, 2014.
In the meeting participated all the members of the Project Board and the invited external scientific experts.
Strategic Project Meeting
Stand up discussion around coffee
The Advisory Board of Potential Users (ABPU) meeting was organized by FCUL, the coordinator of the QTLeap Project and it took place in Lisbon, March 14, 2014.
In the meeting participated all the advisory board members, the invited external scientific experts and all the project members.
This Strategic Project Meeting aimed to endorse the interaction between the Advisory Board of Potential Users (ABPU) and the consortium. The members of the Advisory Board of Potential Users are the industrial participants in the project, and they will have a first hand, inside view of the project and of its results, and will be in the first row of the potential users that will take the lead to exploit their business potential.
Strategic Project Meeting
Networking around coffee
António Branco, Petya Osenova, Kiril Simov and Jan Hajič, from the QTLeap Project, participated in the PARSEME (PARSing and Multi-word Expressions) 2nd general meeting, hosted by the Institute for Language and Speech Processing of the Athena Research Center. This meeting took place in Athens, Greece, from 10 to 11 March 2014.
António Branco was Invited Speaker at the 12th Workshop of TLT – Treebanks and Linguistic Theories that took place in Sofia, Bulgaria in December 13-14, 2013, hosted by the BulTreeBank Group.
The workshop took place at the Sofia University “St. Kliment Ohridski”
The QTLeap Kickoff Meeting took place in University of Lisbon, on November 5, 2013. The meeting was attended by all the partners’ representatives. European Commission was represented by Mr. Wolfgang Treinen who participated in the meeting via teleconference in the afternoon session.
Partners in the Kickoff Meeting