Monthly Archives: July 2013

Workshop on Agile Software Management for Research Environments

We are glad to share with you the materials of the Workshop on Agile Software Managment for Research Environments created within the Abu-MaTran project as the first of the outreach activities planned.

This workshop took place in April 2013 at Dublin City University (Ireland) and aims at defining basic and feasible best practices involving some of the key topics related to software management.

Both for new or senior researchers involved in project managing, this workshop can be a good starting point to make research results turn into successful, well organised and visible projects.

We are making available:

– the workshop guide: abumatran-deliverable-2.1-workshop-guide
– the workshop slides: abumatran-deliverable-2.1-worshop-slides
– a link to the talk given by Dr. Lluís Padró during the workshop about Freeling

All materials are distributed with free/open-source licenses. For a copy of the original files, just e-mail us as indicated in the PDF files.

If you use this materials, please, let us know. We will be glad to have some feedback from you.

Talk by Gema Ramírez Sánchez at ILSP

Project researcher Gema Ramírez Sánchez is giving an invited talk on July 17th at ILSP/Athena Research Centre

The Apertium Machine Translation platform and Prompsit: opportunities for research and business.

Abstract 
Apertium (http://www.apertium.org) is a free/open-source rule-based machine translation platform that provides tools for managing the linguistic data necessary to build a machine translation system for a given language pair, and linguistic data for a growing number of language pairs. The participation of engineers and linguists from Prompsit Language Engineering SA (http://www.prompsit.com) in the development of Apertium since its beginning 8 years ago, has made Prompsit an expert in this technology and its potential applications. We are currently developing and helping the development of more than 20 pairs of languages inside the Apertium platform. We are also developing hybrid systems combining rule-based approaches to MT with translation memories and statistical and example based MT systems, and we have been involved in the development of the parallel corpora crawler Bitextor (http://bitextor.sourceforge.net/). While machine translation has been the core business of Prompsit, an increasing demand on related technologies for named-entity or opinion classification has also brought new development and business opportunities to the company, which sees itself as a vehicle of marketing applied-research results as services.

Short-bio
Gema Ramírez-Sánchez is a translator, a computational linguist and the CEO of Prompsit Language Engineering, a spin-off company which was created inside the Transducens Research Group at the University of Alicante (http://transducens.dlsi.ua.es/) in 2006. Prompsit specialises in language technologies, mainly machine translation, information extraction and sentiment analysis. Gema is visiting ILSP in the framework of Abu-MaTran (Automatic Building of Machine Translation, https://www.abumatran.eu/), an FP7 Marie Curie Industry-Academia Partnerships and Pathways project that aims to enhance industry-academia cooperation in developing and exploiting Machine Translation technologies and resources. Prompsit and ILSP plan to exchange expertise in technologies for automatic acquisition of parallel and monolingual resources from the web.

Abu-MaTran translator is on!

Abu-MaTran releases an open-source machine translation system for Croatian on the day Croatia joins the EU

[Croatian version below. Hrvatska inačica niže]

abumatran-translator-1st-july-2013

New tool will reduce time and costs associated with translation between English and Croatian

The Abu-MaTran project consortium is proud to announce the release of a web-based machine translation system between English and Croatian on the day Croatia is officially joining the EU. To the best of the project partners’ knowledge, the Abu-MaTran system is the first translator for these languages based on free/open-source technologies. The tool will help to reduce the time and costs associated with translation between English and Croatian, which today becomes the 24th language of the EU.

This release marks the first milestone of the EU-funded Abu-MaTran project, just 6 months after the launch of the academia-industry consortium in January 2013. Such quick development has been possible thanks to the mature status of open-source tools for machine translation and the public availability of necessary data sets (texts, dictionaries and rules) for these languages.

The system will serve as a useful and functional tool to aid translation between English and Croatian. It is a baseline system that demonstrates the performance that can be achieved by assembling current state-of-the art technologies and data. During the remaining 3.5 years of the project, the consortium partners will work to improve the system by incorporating pioneering research technologies, high-quality domain-adapted data as well as extending the system to related languages.

Abu-MaTran is an EU Marie Curie project that seeks to enhance industry-academia cooperation as a key aspect to tackle one of Europe’s biggest challenges: multilinguality. The consortium intends to increase the hitherto low industrial adoption of machine translation by identifying crucial cutting-edge research techniques and preparing them for commercial exploitation. Simultaneously, Abu-MaTran aims to transfer back to academia the know-how of industry regarding management, processes and productisation in order to make research results more robust and marketable.

Partners include Dublin City University (coordinator), Prompsit Language Engineering, Universitat d’Alacant, University of Zagreb, Faculty of Humanities and Social Sciences and Athena Research and Innovation Center (Institute for Language and Speech Processing).

Try the English-Croatian translator at http://translator.abumatran.eu

Abu-MaTran objavljuje sustav za strojno prevođenje temeljen na otvorenom kodu za hrvatski jezik na dan pristupa Hrvatske EU

abumatran-translator-1st-july-2013

Novi će alat skratiti vrijeme i sniziti troškove povezane s prevođenjem između engleskog i hrvatskog jezika

Konzorcij projekta Abu-MaTran ponosno predstavlja puštanje u rad mrežnog sustava za strojno prevođenje između engleskog i hrvatskog jezika na dan kada Hrvatska službeno pristupa EU. Prema saznanjima projektnih partnera, Abu-MaTran sustav je prvi prevoditelj za ta dva jezika temeljen na tehnologijama slobodnog i otvorenog koda. Alat će pomoći u tome da se skrati vrijeme i snize troškovi povezani s prevođenjem između engleskog i hrvatskog jezika koji danas postaje 24. jezik EU.

Ova objava predstavlja prvu ključnu točku u projektu Abu-MaTran financiranom od strane EU, u kojem sudjeluju partneri iz znanosti i gospodarstva, i to samo 6 mjeseci od pokretanja konzorcija u siječnju 2013. godine. Ovako brz razvoj sustava bio je moguć zbog zrelosti alata otvorenog koda za strojno prevođenje te javne dostupnosti potrebnih skupova podataka (tekstova, rječnika i pravila) za dotične jezike.

Sustav će služiti kao koristan alat u prevođenju između engleskog i hrvatskog jezika, a prikazuje rezultate koje je moguće postići korištenjem suvremenih tehnologija i podataka. Tijekom preostale tri i pol godine projekta partneri konzorcija će raditi na poboljšanju sustava uključivanjem tehnologija iz novih istraživanja, visokokvalitetnih podataka prilagođenih pojedinim domenama te proširenjem sustava na bliske jezike.

Abu-MaTran je EU Marie Curie projekt kojemu je cilj unaprijediti suradnju između gospodarstva i znanosti što je ključno u rješavanju jednog od najvećih izazova Europe: višejezičnosti. Namjera je konzorcija povećati dosad nizak stupanj korištenja strojnog prevođenja u tržišne svrhe tako što će prepoznati presudne tehnike u posljednjim znanstvenim dostignućima i pripremiti ih za komercijalnu upotrebu. Istovremeno, cilj projekta Abu-MaTran je prenijeti u znanost praktična znanja tržišnog sektora o upravljanju, procesima i oblikovanju proizvoda kako bi rezultati znanstvenog rada bili robusniji te time privlačniji tržištu.

Partneri projekta su sljedeći: Dublin City University (koordinator), Prompsit Language Engineering, Universitat d’Alacant, Sveučilište u Zagrebu, Filozofski fakultet and http://translator.abumatran.eu