All posts tagged Machine Translation

September 15, 2011

SemLab talk on SMT and News Sentiment Analysis at Multilingualweb meeting (W3C)

SemLab’s project director Thomas Dohmen will give a presentation on the benefits and problems of using Statistical Machine Translation for News Sentiment Analysis at the Multilingualweb workshop in Limerick, Ireland.

SemLab is a partner in the LetsMT project for machine translation of smaller European languages, such as Latvian, Danish, Dutch and Croatian. The resulting translation engines of the LetsMT project are showcased on SemLab’s Newssentiment.eu website once completed.

For more information, please visit:
Multilingual web
LetsMT
Newssentiment.eu

 

June 14, 2011

MASS Proposal passes first FP7 stage

SemLab has passed the first stage with their MASS proposal for the FP7 call on Digital Content and Languages, a funding scheme of the European Commission.

In this proposal, SemLab and innovative European partner organisations propose to create a functioning multimedia sentiment analysis platform that incorporates automatic semantic analysis of customer feedback from a wide variety of web-based multimedia, and automatic translation of small and large European languages.

In order to understand social customers, companies must understand the (often unknown) languages of their customers. Today’s existing social CRM solutions are monolingual and are not suitable for following consumers’ responses across Europe. By incorporating automatic translation for larger and smaller European languages into the system, MASS can step beyond the current state of the art and deliver a database of social media information and a system that can determine sentiment from different language groups, cultures, and countries from these data.

This will enable companies to tap into a large base of social customer sentiment, and use this in their CRM solutions. By using the latest semantic analysis technologies, the sentiment portrayed in social media can be extracted. Furthermore, smart semantic tools are used to interpret specific web language, such as shortened words or expressions, emoticons, new ‘internet’ language, etc.

 

May 12, 2011

Positive evaluation of Let’sMT! project’s 1st year

SemLab and the other partners of the Let’sMT project have met in Luxembourg to present the results of the first year of the project.

Presentations were given for each of the work packages as well as a live demo of the already operational translation system, the translation widget and browser plugin.

The preliminary response from the evaluators was very positive and the project will continue as planned.

 

February 15, 2011

Final SOLIM project meeting in Istanbul, Turkey

 

The SOLIM project (Spatial Ontology Language for multimedia Information Modelling) is nearing completion. This week the project partners will meet in Istanbul to discuss the project results and possible future projects.

During the project we have developed two proof of concept applications that show the advantages of using spatial ontology languages for Image Search and Machine Translation technologies.

 

August 20, 2010

SOLIM project in last phase

The Solim project for spatial reasoning has entered the last phase, in which the technology developed will be implemented into two POC applications for evaluation. The two applications that will utilise the spatial reasoning technology are Picsearch’s Image Searching system and Tilde’s Machine Translation system.

 

October 12, 2009

Semlab in Machine Translation project: Let’sMT!

Letsmt
Semlab had joined an international consortium of researchers, linguistic experts and software developers to create a system that automatically translates relatively small languages.

The focus in the project lies with small languages such as Latvian, Lithuanian, Croatian, etc. and one of the main methods that will be used is Statistical Machine Translation (SMT). SMT systems are built by analyzing huge volumes of parallel corpus and learning translation models from these data. This is particularly tricky with smaller languages since the available corpus data is much smaller.

The project aims at a.o. developing a widget or browser add-on that translates such small languages and is currently under evaluation at the European union.

For more information: Let’sMT! Project website

 

October 5, 2009

Solim project meeting in Riga

solim_logo_klein
The Solim project’s spatial reasoning system has been in development for one year and Semlab, Picsearch and AGMlab will meet their partner Tilde at their office in Riga this week to discuss the project’s future.

Last year, the project members have determined the application’s requirements and designed the language, reasoner, translation and image search functionality of the proof of concept (PoC) application. In the coming year, these separate elements will be implemented in order to build the final application.

For more information: Solim Project website