META-NET Workshop: Machine Translation and Multimodal Contexts Special session in ICANN 2012, Lausanne, Switzerland, Sep 11-14, 2012
DL for presentation proposals: 12th of August, 2012
Machine Translation can be considered to be one of the most challenging tasks computer science has ever taken. Statistical methods have been increasingly successful in providing efficiently MT solutions for many language pairs. However, there is a lot of room for improvement regarding the quality of translations. One central reason for the failures is that current systems take the context into account only in a limited manner. This is particularly true for multimodal contexts.
This META-NET workshop is organized to foster exchange of ideas and results which combine machine translation and multilinguality challenges with multimodal processing and adaptive systems.
The notion of context is meant to be understood broadly, including other modalities like vision in addition to the textual contexts.
Therefore, we expect participants, e.g., from machine translation, machine learning, vision research and cognitive systems communities.
Relevant topics include:
* Specification and standardization of contextual information
* Using textual context in Machine Translation
* Visually informed Machine Translation:
e.g. subtitle translation, personal mobile assistants,
multilingual robotics,semantics of movement
* Speech and sound content in multilingual systems:
e.g. context-aware speech-to-speech Machine Translation,
speaker-based speech adaptation,
recipient-directed speech and translation adaptation,
author and content based style adaptation
* Use of situational and location data
* Machine Translation of colloquial language in rapidly changing
* Effect of document type, genre, domain of application and medium on
* Semantics versus pragmatics in multimodal and multilingual systems
* Tools and corpora for context-aware Machine Translation
* Representing Machine Translation problems in a suitable form for
Neural Network and Machine Learning researchers
* Machine Learning methods and tools for context-aware multilingual
* Studies of human translation performance (with and without context)
* Brain research results on context processing
* Learning correspondences between several languages:
e.g. Meta-learning and multi-task learning for multilingual systems
=== Submission of presentation proposals ===
We invite presentations in the domain of the workshop related to
* completed research results,
* work in progress,
* strategic visions and plans,
* promising ideas, and
* META-NET Challenge: Context in Machine Translation
Please, send your presentation proposal to timo dot honkela at aalto dot fi and jaakko dot j dot vayrynen at aalto dot fi by 12th of August, 2012. Notification will be sent by 22nd of August.
A presentation proposal consists of a title, presenter, and an abstract.
=== Program and advisory committee ===
Nicola Cancedda, Xerox, France
Jussi Karlgren, SICS, Sweden
Philipp Koehn, University of Edinburgh, UK Markus Koskela, Aalto University, Finland Mikko Kurimo, Aalto University, Finland Krister Lindén, University of Helsinki, Finland Tapani Raiko, Aalto University, Finland Georg Rehm, DFKI, Germany Francois Yvon, LIMSI, France Pierre Zweigenbaum, LIMSI, France
The workshop program and schedule will be published closer to the workshop.
=== Workshop chairs and contact information ===
Timo Honkela, Aalto University (http://users.ics.tkk.fi/tho/) Jaakko Väyrynen, Aalto University (http://users.ics.tkk.fi/jjvayryn/)