TSM @ AMTA
The TSM workshop will take place at the conference hotel of AMTA 2012. Detailed information will follow soon.
You may register for TSM via the AMTA 2012 official registration website. As soon as registration is open, we will give further information.
- Full Paper submission deadline: July 31st
- Acceptance/Rejection: August 25th
- Camera Ready Paper: September 1st
- Workshop: November 1st
Paper submission is now open (via START) under: https://www.softconf.com/amta2012/TSM2012/cgi-bin/scmd.cgi?scmd=basicSubmit
Papers should not be longer than 10 pages and should be in pdf format. Style files (Latex and MS Word) are available here.
Further information will be published soon
- Chairs: Satoshi Nakamura (NAIST, Japan) and Alex Waibel (KIT, Germany)
- Program Chairs: Graham Neubig (NAIST, Japan), Sebastian Stüker (KIT, Germany), and Joy Ying Zhang (CMU-SV, USA)
- Publicity Chair: Margit Rödder (KIT, Germany)
Call for Participation
During the last couple of years, user generated content on the World Wide Web has increased significantly. Users post status updates, comments, news and observations on services like Twitter, they communicate with networks of friends through web pages like Facebook, and they produce and publish audio and audio-visual content, such as comments, lectures or entertainment in the form of videos on platforms such as YouTube, and as Podcasts, e.g., via iTunes. Nowadays, users do not publish content mainly in English anymore, instead they publish in a multitude of languages. This means that due to the language barrier, many users cannot access all available content. The use of machine and speech translation technology can help bridge the language barrier in these situations.
However, in order to automatically translate these new domains we expect several obstacles to be overcome:
- Speech recognition and translation systems need to be able to rapidly adapt to rapidly changing topics as user generated content shifts in focus and topic.
- Text and speech in social media will be extremely noisy, ungrammatical and will not adhere to conventional rules, instead following its own, continuously changing conventions.
At the same time we expect to discover new possibilities to exploit social media content for improving speech recognition and translation systems in an opportunistic way, e.g., by finding and utilizing parallel corpora in multiple languages addressing the same topics, or by utilizing additional meta-information available to the content, such as tags, comments, key-word lists. Also, the network structure in social media could provide valuable information in translating its content.
The goal of this workshop is to bring together researchers in the area of machine and speech translation in order to discuss the challenges brought up by the content of social media, such as Facebook, Twitter, YouTube videos and podcasts.
We expect participants to submit discussion papers that argue for new research and techniques necessary for dealing with machine and speech translation in the domain outlined above, as well as papers presenting results of related and potentially preliminary research that is breaking new ground.
Karlsruhe Institute of Technology
D – 76131 Karlsruhe
Phone: +49 721 608 48676
Fax: +49 721 60 77 21
Mail: margit roedder∂kit edu