News

The first seminar in the 2012 Dublin Computational Linguistic Research Seminars (DCLRS) will be held on Friday, 3rd February 2012 at 4pm in L2.21 in the
School of Computing, Dublin City University.

DCLRS provides the opportunity for computational linguistic researchers to share their current work with Dublin-based colleagues and friends. The series is presented by researchers at University of Dublin, Trinity College (TCD); University College Dublin (UCD); Dublin City University (DCU); and Dublin Insitute of Technology (DIT). The 2012 Series is hosted by DCU.

Details of the 3rd February seminar, as follows:

Lingle : Creating the next generation of online language learning tools

Presented by:

Dr. John Kelleher (Applied Intelligence Research Center, Dublin Institute of Technology)
Dr. Brian Mac Namee (Applied Intelligence Research Center, Dublin Institute of Technology)
Dr. Noel Fitzpatrick (School of Art, Design and Printing, Dublin Institute of Technology)

Lingle (www.lingleonline.com) is a commercial web-based service designed for teachers of English as a second language. Lingle helps teachers automatically find and build complete lesson plans andm materials from authentic real news sources.

Apart from the standard keyword and topic searches Lingle also allows teachers to search for appropriate content by difficulty level (i.e., Pre-Intermediate, Intermediate, Upper-Intermediate, Advanced) and by grammar point (e.g., first-conditional tense, etc.). The talk will present the Lingle system and some of the research it is based upon. It will also describe the experiences of the presenters in taking a research idea and developing a commercial product.

Dublin City of Science logo

ESOF2012 Dublin City of Science was officially launched today (26 January 2012) at the Convention Centre Dublin. The Centre for Next Generation Localisation (CNGL) features on the Public Engagement Programme with the national finals of the All Ireland Linguistics Olympiad in March and the CNGL Localisation Innovation Showcase this autumn. Details of both events will be published shortly.

The Dublin City of Science 2012 festival is a year-long celebration of science with over 160 events that will awaken your curiosity and open your mind to science. Join us in an exciting adventure of exploration that combines the best of Irish culture, arts and science like you have never seen before.

The programme of science themed events across Dublin and Ireland consists of photographic and art exhibitions, several theatre pieces, film festivals, tours, trails and treasure hunts, street performances, science buskers, large-scale interactive installations, experiments, science carnivals, public talks, debates, workshops and even a fashion show.

We look forward to what promises to be an exciting year of science and technology!

Professor Josef Van Genabith pictured with Mr Sean Sherlock TD, Minister for Research and Innovation

Prof Josef Van Genabith, Director of the Centre for Next Generation Localisation (CNGL), with Mr Sean Sherlock T.D., Minister for Research and Innovation, launching CNGL's new localisation careers guide

Minister Sherlock supports call for more skilled graduates in language and technology to maintain Ireland’s world leadership position in multi-billion euro localisation sector

Seán Sherlock T.D., Minister for Research and Innovation today (24 January 2012) launched a guide which urges secondary students to consider language and technology at third level for a potential career in the localisation sector.

The Guide, produced by the Centre for Next Generation Localisation, an academia-industry research consortium supported by the Department of Jobs, Enterprise and Innovation through Science Foundation Ireland, says that Ireland needs to significantly increase its supply of language and technology graduates if the country is to maintain its leadership position in the multi-million euro localisation and global services sector.

Speaking at today’s launch, Minister Sherlock said that localisation, the process of adapting and personalising digital products, services and content to the needs of global users was an area of growth in Ireland and that opportunities exist for future graduates to build a career in this rapidly growing field.

Minister Sherlock stated “Many of the world’s largest software and web companies co-ordinate their localisation activities in Ireland and the sector contributes an estimated €680 million annually to the Irish economy.”

The Minister added “It is imperative that we highlight the opportunities available to our young people in this sector. We need to encourage and foster their development to ensure that we are teaching the right skills at third-level so employers who are already here, or who are considering a presence here, have a strong supply of high-calibre graduates to help them grow their businesses.”

The CNGL consortium comprises four Irish universities and includes Irish-based multinationals Microsoft, IBM and Symantec, as well as SMEs such as Alchemy Software Development, VistaTEC, Welocalize, and Applied Language Solutions.

The guide highlights the range of career opportunities available in this sector. The guide includes profiles of individuals who have graduated from Irish universities in a range of disciplines who now hold leading roles across the Localisation industry and also lists the University courses across computing, languages, and business that would provide a foundation for a career in this high-growth area.

“At present there are approximately 16,000 people working directly in the localisation sector in Ireland”, explains Dr Páraic Sheridan, Associate Director at CNGL. “Localisation is critical to so many industries, including software, games, financial services and medical devices, in bringing their products to foreign markets, so it really is vital to Ireland’s export-led recovery. It is also an area where Ireland is widely recognised as a world leader in innovation and sees continued growth in jobs, not just in traditional product and service localisation but also in the areas of global multilingual customer care and support.”

By launching a localisation careers guide, CNGL hopes to put localisation firmly on the radars of students, guidance counsellors and parents and to help ensure an adequate number of language, technology and linguistics graduates to support the sector’s continued success.

Dr Fred Hollowood, Research Director at Symantec Corporation’s Shared Engineering Services group in Blanchardstown supports CNGL’s call for students to consider studies in localisation-related fields. “Localisation enables Symantec to deliver products and offer customer support to customers in 40 countries in 22 languages out of our Dublin offices”, says Hollowood. “Although still a relatively young industry, localisation is expanding at a rapid pace. It is ideal for students who have an interest in language and technology, who are interested in the difference between the cultures of the world, and who are seeking a career in a fast growing sector.”

Director-General of Science Foundation Ireland (SFI), Professor Mark Ferguson, welcomed the publication of the guide. “A key goal of Science Foundation Ireland is to ensure that Ireland has a pipeline of skilled graduates who can pursue next generation science careers. This guide to careers in the localisation industry reinforces the fact that science offers a diverse range of roles that are rewarding, in strong supply and central to our economic growth”, says Professor Ferguson.

CNGL’s new localisation careers guide is being distributed to all secondary schools in Ireland this week. It features profiles from industry professionals, insights into the variety of job roles on offer, and details of relevant third level courses. The guide can also be downloaded at www.cngl.ie/careers.

DCU@FIRE 2011: SMS-Based FAQ Retrieval

Speaker: Johannes Leveling
Date: Wednesday January 25th 2012
Time: 4pm
Location: L2.21, School of Computing, DCU

Abstract:
This presentation gives an overview of the DCU participation in the SMS-based FAQ Retrieval task at FIRE, the Forum for Information Retrieval Evaluation. The objective in the SMS-based FAQ retrieval task is to find answers in a collection of frequently asked questions (FAQ) given a SMS question in "text-speak".

DCU submitted experimental runs for the monolingual English subtask. The DCU approach to this problem consists of first transforming the noisy SMS queries into a normalised, corrected form. The normalised queries are then used to retrieve a ranked list of FAQ results by combining the results from three slightly different retrieval mechanisms. Finally, using information from the retrieval results,
out-of-domain (OOD) queries are identified and tagged.

The results of our best run on the final test set are the best among 13 participating groups. We retrieved results for 70% in-domain queries correctly, identified 85.6% out-of-domain queries correctly, and obtained an MMR score of 0.896.

View further details of the NCLT/CNGL Seminar Series 2011/2012

CNGL was on the SFI stand at the BT Young Scientist & Technology Exhibition at the RDS, Dublin on Thursday 12th January 2012.

The Education and Outreach team were joined by John Moran from CNGL TCD to meet the primary and secondary school students. John showed the students his ipad German cloze test app. The students also completed All Ireland Linguistics Olympiad (AILO) puzzles on the stand and were given information on undergraduate courses at the CNGL universities.

Minister for Research and Innovation, Seán Sherlock T.D. visited CNGL at the SFI stand. He met students doing the Olympiad puzzles and browsed the new CNGL Localisation Careers guide.

John Moran at Young Scientist 2012 Minister for Research and Innovation, Seán Sherlock T.D. browses the CNGL Localisation Careers Guide with Laura Grehan of CNGL
John Moran at Young Scientist 2012 Minister for Research and Innovation, Seán Sherlock T.D. browses the CNGL Localisation Careers Guide with Laura Grehan of CNGL

The registration deadline for the 2012 season of CNGL's popular All Ireland Linguistics Olympiad is fast approaching!

Registration will close on Friday, 20th January 2012. Don't miss out. Register now at www.cngl.ie/ailo

The All Ireland Linguistics Olympiad (AILO) is a secondary schools competition that challenges students to develop their own strategies for solving problems in unfamiliar languages.

Could you decipher Sanskrit poetry or decode Faroese expressions by using just your ingenuity and logic? AILO challenges you to solve some of the world's toughest puzzles in language, logic and linguistics.

Join the record 700 students from 60 schools who have already signed up for the challenge this year. You could be one of four students selected to represent Ireland at the International Linguistics Olympiad in Slovenia in July 2012.

To see if you have a knack for problem-solving, why not try out some sample puzzles yourself at www.cngl.ie/ailo

Detecting Grammatical Errors with Treebank-Induced, Probabilistic Parsers

Speaker: Joachim Wagner
Date: Wednesday January 18th 2012
Time: 4pm
Location: L2.21, School of Computing, DCU

Abstract:
At first glance, treebank-induced grammars seem to be unsuitable for grammar checking as they massively over-generate and fail to reject ungrammatical input due to their high robustness. In this talk, I give an overview of my research on applying such grammars to automatically judge the grammaticality of an input string. I show evidence that grammaticality is reflected in the generative probability of the best parse, discuss training and evaluation of classifiers and present results for 9 selected methods, including baseline XLE grammar and n-gram methods and machine learning-based methods.

View further details of the NCLT/CNGL Seminar Series 2011/2012

A Date for your Diary!

CNGL's Localisation Research Centre at the University of Limerick has confirmed that the LRC XVII Conference will be held on 19th - 21st September 2012 in Limerick.

The annual LRC Conference brings together the innovators and movers of 21st century localisation strategies to discuss how their organisations can become a leader in next generation localisation.

More details to follow.

The below article by Grainne Rothery featured on the website of Business & Leadership on 5th January 2012. View the original article.

---

US-headquartered language translation company Cetra is establishing its European Service Centre in Limerick with the support of Government through IDA Ireland.

The company plans to create 20 jobs in Limerick to better serve its existing European clients and further grow its EMEA client base.

Irish based employees will include management, administration, project managers and sales staff. The company, which is set to start recruiting for initial positions immediately, will initially locate in the Limerick Institute of Technology Acceleration Centre.

Cetra offers translation, localisation and interpretation services to the market research, legal, and life science industries using a global network of professional linguists.

The announcement was made this morning by the The Minister for Jobs, Enterprise and Innovation Richard Bruton TD.

“This is a high-growth international company which has chosen to locate in Ireland with the creation of jobs due to the graduate output of one of our universities,” he said. “With the right policies we can replicate this announcement across the country over the coming years.”

Dr Jiri Stejskal, president and CEO of Cetra said the company was attracted to Limerick primarily due to the graduate output of the University of Limerick and the ability to collaborate with the Localisation Research Centre in UL to develop a technology localisation offering. “We’re excited about expanding our business into EMEA from our Limerick base,” he said.

700 Students to test their Minds with AILO 2012

Students from Mount Temple Comprehensive School pictured at the launch of the 2012 season of AILO

The popularity of CNGL’s All Ireland Linguistics Olympiad appears to know no bounds. A rigorous marketing campaign, launched in September, has produced a surge in enquiries and registrations for the novel secondary schools competition.

To date more than 700 students from 58 schools across 22 counties have registered for the 2012 season. This represents a 30% increase over 2011 figures.

“We are delighted that so many students will this year have the opportunity to sample the field of linguistics for themselves and to develop their own strategies for solving problems in fascinating languages from around the globe”, says CNGL Education & Outreach Manager, Cara Greene.

As well as appealing to language teachers, the competition is now attracting significant interest from guidance counsellors in schools, who appreciate the competition’s potential for honing students’ soft skills in areas such as problem solving and team work. At the IBEC/Institute of Guidance Counsellors Careers Day in Cork last month, the AILO stand was inundated with information requests relating to the competition.

Participating schools will host qualifying rounds in February, with the top 100 students selected to compete at the national finals of the competition at Dublin City University in March. The four best performers at the national finals will be selected to represent Ireland at the International Linguistics Olympiad in Slovenia in June 2012.

The deadline for registration for the 2012 season of the All Ireland Linguistics Olympiad is 20th January 2012. Register online at www.cngl.ie/ailo

Cover of Localisation Focus Vol 10 Issue 1

We are pleased to announce that Localisation Focus - The International Journal of Localisation, the original, peer reviewed and indexed journal dedicated to localisation, internationalisation, language and related technologies, will be distributed free of charge in electronic format from the current (2011) edition on.

Regular subscribers will still receive their high-quality printed copies with free subscribers receiving a special pdf edition.

Localisation Focus is produced by CNGL staff at the Localisation Research Centre, University of Limerick.

This current edition features two articles by CNGL researchers:

An Argument for Business Process Management in Localisation (David Filip, Eoin Ó Conchúir)

LocConnect: Orchestrating Interoperability in a Service-Oriented Localisation Architecture (Asanka Wasala, Ian O'Keeffe, Reinhard Schaler)

View the 2011 edition of Localisation Focus (pdf)

The Quarter 4 2011 edition of 'CNGL News', the quarterly newsletter of the Centre for Next Generation Localisation (CNGL) is now available.

Download the CNGL Newsletter (pdf - 3.53MB).

Feature articles this quarter include:

* Minister for Jobs, Enterprise and Innovation hails Value of Localisation to Irish Economy
* Looking ahead: Towards Global Content Intelligence
* CNGL Researchers on Path to Commericalisation
* 700 Students to test their Minds with All Ireland Linguistics Olympiad (AILO)
* Action for Global Information Sharing (AGIS'11)

Plus, you can read event reviews, research track updates, details of recent funding awards, and much more.

We hope you enjoy this edition!

The CNGL Team

To keep abreast of news, events and developments at CNGL, subscribe to CNGL News Alerts

Reinhard Schaler, Chris Exton and Solomon Gizaw of CNGL pictured with Hans Fenstermacher of GALA and local farmers

CNGL researchers at the University of Limerick co-organised AGIS'11 —The Action Week for Global Information Sharing—earlier this month in Ethiopia. Here, Reinhard Schäler of UL reviews the conference.

---

AGIS ‘11
Where the streets have no names…

“I’ll show you a place
High on a desert plain
Where the Streets have no name”
U2

We had not expected perfect mobile phone access and broadband internet in Lalibela, about an hour’s flight up north from Ethiopia’s eight million capital, Addis Ababa – even less so in Yemirhane Kristos, an ancient monastic settlement 2,800m high up in the mountains of northern Ethiopia, a good two hours drive from Lalibela on dusty and at times terrifyingly dangerous roads. But – we found what we were looking for: perfect mobile phone connections, even though for many villagers the next re-charging station was a five-hour walk away.

We learnt that Africa is not what most of us ‘westeners’ expect it to be. Or did you know that, over the past decade, six of the world’s ten fastest-growing countries were African; that in eight of the past ten years, Africa has grown faster than East Asia, including Japan; and that with 600 million mobile phone users, Africa has overtaken America and Europe?

It is because of our ignorance of Africa that the hottest continent is still not on the radar of mainstream localisation. This became clear at the recent GALA-and Localization World supported Action for Global Information Sharing event (AGIS ’11), hosted by the UN’s Economic Commission for Africa (UNECA) in Addis Ababa, co-organized by the Localisation Research Centre (LRC) and supported by theCNGL. Have you heard about the excellent and ground-breaking work of the African Network for Localisation (ANLoc); the African Language Technology Network (AfLaT); or the African language Kamusi dictionary project?

Africa is the cradle of humanity, the birthplace of human language and, with close to 3,000 languages, the world’s most diverse linguistic continent. Contrary to the western view of Africa, where many still believe that the former colonial languages, Portuguese, English and French, ‘will do’ at least for business, tens of millions of people just speak languages such as Arabic, Berber, Igbo, Somali, Swahili, Hausa, Amharic, Oromo or Yoruba – and among them millions of local traders and farmers on national commodity exchanges.

Language policy in this highly diverse continent has become a vital issue, with African countries becoming increasingly aware of the value and the opportunities presented to them by their linguistic inheritance. All African languages are considered official languages of the African Union (AU) who declared 2006 as the “Year of African Languages”.

At AGIS ’11 earlier this month, close to 200 delegates gathered in the Ethiopian capital, Addis Ababa, to celebrate the Third Action for Global Information Sharing (AGIS ’11). The UN’s Economic Commission for Africa (ECA) hosted the event, which was opened by the Ethiopian Minister for IT, the Irish Ambassador to Ethiopia, the Director, ICT Science and Technology (ECA), and the Director of the Localisation Research Centre (LRC) at the University of Limerick.

For the first time, AGIS ’11 connected Africa’s language industry, decision makers and researchers with the mainstream localization and translation industries and researchers. Ethiopia’s Minister for IT, in his opening address, stressed the importance of Ethiopia’s two hundred language for the economic development of his country. The Irish Ambassador stressed the important role our country, with a population smaller than that of Addis, sees in the economic and cultural development of Ethiopia, and pledged Ireland’s continued support for a multilingual Africa. As an example for new and exciting developments in Ethiopia, she described Ethiopia’s new fully electronic commodity exchange, where local farmer can now trade their crops in their language largely without costly middlemen, thus increasing their income and that of their communities.

The support of Ireland to AGIS ’11 was acknowledged by many speakers, among them the secretary of the Academy of African Languages. Other speakers suggested replacing the western “P” for personal in PC or PA, with a “C” for community in the most more community-based approach to business and development in the African context.

The event was jointly organized by the United Nations’ ECA (UNECA), CDAC Pune, and the LRC. It was supported by a wide range of African organizations, including the African Network for Localization (ANLoc), the African Language Technology Network (AfLaT), and the Internet Society (Africa); western organizations supporting the event included the world’s largest industry association, the Globalization and Localization Association (GALA), the world’s largest industry event organizer, Localization World, and the Centre for Next Generation Localization (CNGL), with Multilingual Computing as the media sponsors.

Delegates agreed that AGIS ’11 will be the beginning of a number of collaboration initiatives and established the AGIS Network to facilitate information and knowledge exchange. UNECA and the LRC signed a Memorandum of Understanding to support these developments; a further memorandum is planned to be signed with the Academy of African Languages in the near future.

The Chair of the Board of GALA, Hans Fenstermacher, joined LRC and CNGL researchers on this exciting and eye-opening journey into the heart of the world’s hottest continent. As a result, concrete discussions are already on the way to explore joined actions to promote the work of Africa’s vibrant language and localisation communities within GALA, and to involve GALA much closer in the preparations for AGIS ’12.

Mainstream localisation, on occasion, has a tendency to talk to itself. In the meantime, the linguistically most diverse continent on the globe, the cradle of humanity and human language, the continent with the world’s fastest economic growth figures, is taking off. It’s decision time. It’s a question of when (not if) the west will join the African renaissance. The streets of Africa are being named – in languages most of us have not even heard about.

AGIS ’12 will take place in Sri Lanka on 10-11 December 2012. Keep the date!

Note: AGIS ’11 would not have happened without the dedication and support of Solomon Gizaw, a PhD student in ‘Next Generation Localisation’ (LOC) at the Localisation Research Centre of the University of Limerick.

Reinhard Schaler and Chris Exton of CNGL pictured with Irish Ambassador Síle Maguire and Hans Fenstermacher of GALA

Reinhard Schaler and Chris Exton of CNGL pictured with Irish Ambassador Síle Maguire and Hans Fenstermacher of GALA

Title: How can NLP contribute to the medical domain? Some examples.

Date: Wednesday, 23 November; 16:00
Venue: L2.21, School of Computing, DCU
Speaker: Lorraine Goeuriot

Everyone is concerned with health topics. Thus, there is a proliferation of health-related textual data: any kind of author (e.g., patients, students, general practitioners, researchers) can write about any kind of topic (e.g., disease, treatment, drug) in any language. What kind of knowledge can we extract from such an amount of data? And how can this contribute to the medical domain?

In this talk, Lorraine will present two projects:

The first one focuses on the compilation of trilingual comparable corpora on specialised domains, especially on the medical one. Those resources are mainly used to build multilingual terminologies, highly needed in specialised domains.

The second project is an opinion mining system on health-related user-generated content. First focusing on drug-related texts, our lexicon-based system provides
summarised view of opinions expressed on different aspects of the drugs.

View more details of the NCLT/CNGL Seminar Series 2011/2012

Mr Richard Bruton TD, Minister for Jobs, Enterprise and Innovation, addresses attendees at the CNGL Localisation Innovation Showcase

SFI research centre created technology used at 2010 FIFA World Cup to automatically translate tweets into global languages

The Minister for Jobs, Enterprise and Innovation, Richard Bruton TD today (16 November 2011) attended a showcase of work by the Science Foundation Ireland-funded Centre for Next Generation Localisation (CNGL) at Croke Park Convention Centre, Dublin to see at first hand the economic benefits being generated by the Government’s investment in third level research.

Technologies created by the centre include the software used at the 2010 FIFA World Cup to translate tweets into global languages, and the centre is continuing to develop ground-breaking technologies in voice recognition and language translation for use in software and social media across the world.

Localisation is a sector within the technology industry, whereby digital products and services are adapted to the needs of customers in countries around the world. Approximately 15,000 people are employed in the sector in Ireland, which is worth an estimated €680million to the economy per year.

Addressing the representatives of industry and research attending the showcase, Minister Bruton said:
“Over recent years Ireland has built up a very impressive record in scientific research. If we are to get out of our present difficulties, we must now use this base to commercialise more ideas, create more Irish businesses, and attract and embed more multinational investment. Turning good ideas into good jobs: that is what SFI’s research centres and funded teams are about, and that is what I am determined to achieve.

“I am delighted to see that five start-up companies have been created from this research, as well as an impressive number of invention disclosures, licensing agreements and patent applications. We must now build on that record and be highly ambitious about the number of jobs our research can create”.

The CNGL is one of 28 top-class research centres funded by Science Foundation Ireland, and brings 150 researchers together with 10 partner companies in order to commercialise research, foster the development of new and existing Irish companies, and attract new multinational investment. The CNGL Localisation Innovation Showcase demonstrated the scientific and economic impact of work at the €40m academia-industry consortium, while also presenting a selection of technologies that are ripe for commercial exploitation.

CNGL is a dynamic consortium comprising more than 150 researchers at Dublin City University, Trinity College Dublin, University College Dublin and University of Limerick, along with ten industry partner companies. In addition to on-going collaboration with its industry partners, CNGL has engaged with more than 100 companies across Ireland over the last year under the Enterprise Ireland Commercial Development Manager programme. CNGL partner Symantec and new start-up Digital Linguistics today shared their experiences of engaging in university research, shaping the commercial opportunity, and then building products and services to capitalise on the innovation.

Summarising the benefits of attendance at today’s Localisation Innovation Showcase, CNGL’s Commercial Development Manager, Steve Gotz, said: “Our showcases have become particularly fertile grounds for companies looking to solve specific problems or find new product offerings. Today attendees had the opportunity to not only view fundamental scientific advances that will be impacting their businesses in the not-too-distant future, but also a chance to discuss industry challenges and opportunities with some of the leading academic minds in these fields.”

Through strong industry engagement, CNGL is helping to ensure that Ireland remains at the forefront of the multi-billion euro content management and localisation sectors. The impact of research in localisation and global content intelligence is particularly significant in an export-led economy such as Ireland, as localisation enables companies to introduce products to otherwise inaccessible markets.

CNGL’s Localisation Innovation Showcase was presented as part of the Innovation Dublin Festival 2011. For further information on collaboration opportunities with CNGL, see http://www.cngl.ie/collaboration.html

Pictured at the CNGL Localisation Innovation Showcase are Dr Páraic Sheridan, Associate Director, CNGL; Dr Graham Love, Director General, Science Foundation Ireland; Mr Richard Bruton, TD, Minister for Jobs, Enterprise & Innovation; Prof Vincent Wade, Deputy Director, CNGL; and Dr Fred Hollowood,Research Director, SES EMEA, Shared Engineering Services, Symantec Corporation.

Pictured at the CNGL Localisation Innovation Showcase are Dr Páraic Sheridan, Associate Director, CNGL; Dr Graham Love, Director General, Science Foundation Ireland; Mr Richard Bruton, TD, Minister for Jobs, Enterprise & Innovation; Prof Vincent Wade, Deputy Director, CNGL; and Dr Fred Hollowood,Research Director, SES EMEA, Shared Engineering Services, Symantec Corporation

Eva Szekely, a CNGL PhD student based in UCD, overcame 24 other PhD students from CSETs around Ireland to win last night's Thesis in Three competition.

Eva Szekely receiving her award from Dr. Stephen Flinter SFI Eva Szekely receiving her award from head judge Dr. Stephen Flinter (SFI)

As part of the Innovation Dublin festival, and sponsored by SFI, DSE and CPL, 25 PhD students from the 9 SFI-funded Centres for Science, Engineering and Technology (CNGL, CLARITY, SBI, CRANN, BDI, Lero, CTVR, APC, DERI) presented their research in bite-sized chunks - three slides in three minutes. The competition took place in the Mansion House, Dublin with an audience of 300 people.

Eva won first place in the competition for her presentation "Voices that Speak for You", which described her work on infusing emotion in speech synthesis. Arlene O'Neill from CRANN won second prize for her presentation "Building our future: One nano layer at a time" and Sean Fitzpatrick from BDI won third prize for 'How TRIM is your Bowel?' - Marker Protein Expression in Colorectal Cancer". Jason Jensen from CRANN won the audience prize for "Nano-Bio Diagnostics Using Cantilevers".

Well done to all the CNGL students participating in our initial internal round last month. Congratulations to Hanna Bechara ("Recycling Machine Translation") and Wei Li ("Retrieval Rescued by Recommendation"), who also represented CNGL last night and did us proud.

Eva has been invited to deliver her winning presentation at the SFI Summit 2011 next week in Athlone. A big "congratulations" to Eva from everyone at CNGL!

Date:Wednesday, 9 November 2011
Time: 16:00
Location: L2.21, School of Computing, Dublin City University

Speaker: Hala Almaghout

Abstract: Hierarchical Phrase-Based (HPB) Machine Translation (MT) system extracts synchronous Context-Free Grammar (CFG) from a parallel corpus without using any syntactic information. Nonterminals in HPB rules act as placeholders which are replaced by other phrases during decoding.

In the baseline HPB system, there are no syntactic constraints imposed on nonterminal replacement during decoding. Methods such Syntax-Augmented Machine Translation (SAMT) try to constraint phrases allowed to replace nonterminals in HPB rules by labelling them with syntactic labels extracted using phrase-structure grammar. However, the effect of using such constraints is limited because their application does not cover all the levels of the derivation for two reasons.

First, these syntactic constraints are applied on hierarchical rules only and do not include glue grammar rules, which perform monotone phrase concatenation in HPB SMT system. Second, phrases which fail to have a syntactic label do not undergo any syntactic constraint during decoding.

In my current work, I will try to approach these two problems by using Combinatory Categorial Grammar (CCG) to control glue grammar-based phrase concatenation. In addition, I will try to increase the coverage of CCG-based syntactic labels by using composite syntactic labels which consist of two or more CCG categories. This work is still in progress, so I will appreciate the feedback from group members, especially the parsing specialists.

All comments are submitted to the feedback forum in the members area.