Intensive School in Machine Translation
Joint EAMT – SUSU Intensive School in Machine Translation
European Association for Machine Translation and South Ural State University are pleased to announce an exciting co-sponsored event “Intensive School in Machine Translation (ISMT)” which will be held in Chelyabinsk, Russia from 16 to 20 May 2011. The venue of the school is South Ural State University.
The overall goal of ISMT is to provide participants with a reference frame for major areas of study within the field of MT technology while defining critical problems of MT and the most useful methods for their solutions.
The ISMT courses on the state-of-the-art MT aspects will be delivered by internationally renowned researchers from different countries. Click here to see the lecturers.
ISMT will also feature a discussion panel which will enable participants to discuss specific issues in MT with experts in the field. The working languages are English and Russian. A course certificate will be issued upon completion of the course, based on 90% participation.
The primary ISMT target audience is university students, PhD students and young researchers for whom attendance is free of charge. The school is also open for everyone else professionally interested in machine translation. Participants from industry/commerce will be charged a registration fee.
Since there are a limited number of participant slots we do recommend interested parties to register as early as possible. Assistance with accommodation is offered once you have registered.
| Important Dates |
| 15 January 2011 |
|
Registration opens Click here to register |
| 15 April 2011 |
|
Registration deadline |
| 16-20 May 2011 |
|
Intensive School in Machine Translation |
Contact: org@mtschool-susu.info
The ISMT courses on the state-of-the-art MT aspects will be delivered by internationally renowned researchers from different countries.
The programme includes:
Introduction to European Asociation for Machine Translation (EAMT) and other (AMTA, AAMT) chapters of International Association for Machine Translation: The role of the EAMT community in MT research and development (R&D), its organizational structure, history and major events
Introduction to MT issues
- Overview of major MT technologies
Why MT is difficult, but not totally impossible. Linguistic problems in MT
Linguistic knowledge elicitation bottleneck
Problems in empirical approaches
Possible solutions
MT for translators and language learners
- Translation memory and other tools for translators
- Use of free online MT(including its use in language teaching)
MT Research and Development
Rule-Based Machine Translation (RBMT)
- Major RBMT architectures: Direct, Transfer and Interligua/Knowledge-Based.
- Knowledge acquisition for RBMT: grammar, lexicons & ontologies.
- Domain use in RBMT
- The KANT Knowledge-Based system
- The AutoTrans Transfer system
Empirical MT
Example-Based Machine Translation (EBMT)
- Problems of EBMT & Generalized Templates
- EBMT System Development
Statistical Machine Translation (SMT)
- Introduction to probability, corpora, data modelling
- Why SMT is hard to understand
- Corpus-based language and translation models.
- Alignment Methods: word-based, character-based etc.
- Probabilistic Word (and Phrase) Models.
- Current research in SMT :Adding syntax back in with statistical learning
Hybrid and Multi-Engine MT (MEMT)
- Domain Analysis and System Specification
- Analyzer & Generator Development
- System combination approaches for MEMT
MT for Speech translation
- Speech recognition and speech synthesis
- Speech to speech translation
- Text to speech translation
- Speech to text translation
- User interface issues for Speech Translation
- Speech MT in medical domains
Future directions for MT research |
|
