Workshop Description

Mobile devices, such as ultra-mobile PCs, personal digital assistants, and smart phones have many unique characteristics that make them both highly desirable as well as difficult to use. On the positive side, they are small, convenient, personalizable, and provide an anytime-anywhere communication capability. Conversely, they have limited input and output capabilities, limited bandwidth, limited memory, and restricted processing power.  

The purpose of this workshop is to provide a forum for discussing the challenges in natural and spoken language processing and systems research that are unique to this domain. We argue that mobile devices not only provide an ideal opportunity for language processing applications but also offer new challenges for NLP and spoken language understanding research.

For instance, mobile devices are beginning to integrate sensors (most commonly for location detection through GPS, Global Positioning Systems) that can be exploited by context/location aware NLP systems; another interesting research direction is the use of information from multiple devices for “distributed” language modeling and inference. To give some concrete examples, knowing the type of web queries made from nearby devices or from a specific location or a specific ‘context’ can be combined for various applications and could potentially improve information retrieval results. Learned language models could be transferred from device to device, propagating and updating the language models continuously and in a decentralized manner.  

Processing and memory limitations incurred by executing NLP and speech recognition on small devices need to be addressed. Some applications and practical considerations may require a client/server or distributed architecture: what are the implications for language processing systems in using such architectures? 

The limitation of the input and output channels necessitates typing on increasingly smaller keyboards which is quite difficult, and similarly reading on small displays is challenging. Speech interfaces for dictation or for understanding navigation commands and/or language models for typing suggestions would enhance the input channel, while NLP systems for text classification, summarization and information extraction would be helpful for the output channel.  Speech interfaces, language generation and dialog systems would provide a natural way to interact with mobile devices.

Furthermore , the growing market of cell phones in developing regions can be used for delivering applications in the areas of health, education and economic growth to rural communities.  Some of the challenges in this area are the limited literacy, the many languages and dialects spoken and the networking infrastructure.

The goal of this one day workshop is to provide a forum to allow both industrial and academic researchers to share their experiences and visions, to present results, compare systems, exchange ideas and formulate common goals.

   

Workshop Program & Accepted Papers 

8:45	Opening Remarks
9:00	Invited talk: Dr. Lisa Stifelman, Principal User Experience Manager at Tellme/Microsoft. Say it and See it! Applying User-Centered Design to Mobile and Multimodal Search
10:00	A Multimodal Home Entertainment Interface via a Mobile Device; Alexander Gruenstein et  al. MIT Computer Science and Artificial Intelligence Laboratory.
10:30	Break
11:00	A Wearable Headset Speech-to-Speech Translation System; Kriste Krstovski et al.  Speech and Language Processing Department, BBN Technologies.
11:25	Information extraction using finite state automata and syllable n-grams in a mobile environment. Choong-Nyoung Seon et al.  Computer Science and Engineering, Sogang University, Seoul, Korea.
11:50	Small Statistical Models by Random Feature Mixing; Kuzman Ganchev and Mark Dredze. Department of Computer and Information Science, University of Pennsylvania, Philadelphia, PA.
12:15	Lunch
1:15	Mixture Pruning and Roughening for Scalable Acoustic Models; David Huggins-Daines and Alexander I. Rudnicky. Language Technologies Institute, Carnegie Mellon University, Pittsburgh.
1:40	Assistive Mobile Communication Support. Sonya Nikolova and Xiaojuan Ma, Princeton University, Princeton.
2:05	A Distributed Database for Mobile NLP Applications. Petr Homola, Institute of Formal and Applied Linguistics, Charles University, Prague, Czech Republic.
2:30	Demos & Posters
3:30	Break
4:00 5:00	Panel Session

 

Invited Talk  

Say it and See it! Applying User-Centered Design to Mobile and Multimodal Search

Dr. Lisa Stifelman
Principal User Experience Manager at Tellme/Microsoft

Bio

In 1999, Lisa was recruited by Mike McCue to join a new startup, Tellme. At Tellme, Lisa led the design of the consumer voice portal, 1 -800-555-TELL, and went on to establish the Tellme User Experience team. Lisa has now been designing speech interfaces for more than 15 years. She grew up on speech at the MIT Media Lab, creating mobile speech designs like VoiceNotes and the AudioNotebook (a device linking paper notes to recorded audio). Lisa’s also worked at AT&T, the Apple Human Interface Group and Interval Research. She holds an M.S. and Ph.D. from MIT, and a B.S. in Human Factors Engineering from Tufts University.

Panel

Susan Boyce, Tellme

Ken Church, Microsoft Research

Michael Johnston, AT&T

Tim Paek, Microsoft Research

Mike Phillips, Vlingo

Barbara Rosario, Intel Research

Noah Smith, CMU

Workshop Chairs

Barbara Rosario   Intel Research
Tim Paek  Microsoft Research

Program Committee

Alex  Acero (Microsoft Research)
Alan Black (CMU)
Dilek Hakkani Tur (ICSI)
Marti Hearst (iSchool, UC Berkeley)
Michael Johnston (AT&T)
Maryam Kamvar (Google & Columbia University)
Kevin Knight (USC/Information Sciences Institute)
Julian Kupiec (Google)
Dekang Lin (University of Alberta, Canada)
Maryam Mahdaviani (University of British Columbia, Canada)
Wolfgang Minker (University of Ulm, Germany)
Noah Smith (CMU)
Bo Thiesson (Microsoft Research)
Gokhan Tur  (SRI)
Fuliang Weng ( Bosch)
Thomas Zheng  (Tsinghua University)
Geoffrey Zweig (Microsoft Research)

Registration 

Information on registration is provided at the ACL-2008 website

 
Contact

For questions about the workshop, please contact Barbara Rosario (barbara.rosario@intel.com)