Workshop on Advances in Text Input Methods (WTIM 2011)

Workshop proceedings are now available.

Keynote Speech by Toshiyuki Masui, Keio University, "Toward a Universal Text Input Method for the Ubiquitous Computing Age" (slides)

Speaker Bio: Toshiyuki Masui is Professor of Faculty of Environment and Information Studies, Keio University, Japan. Before joining Keio University, he worked at many IT companies including Fujitsu, Sharp, Sony, and Apple.

For many years, he has been working on improvements of the user interface of computers, mobile phones, and information appliances in the coming age. In 2001, he developed a predictive Japanese text input system "POBox" for Sony's mobile phones, and the technology is now the standard input method on almost all the Japanese mobile phones. In 2006, he was headhunted from Apple for the development of iPhone's Japanese text input system, and worked there for two years. His current research interest includes text input techniques, information visualization, and low-cost ubiquitous computing. He earned a Doctorate of Engineering from the University of Tokyo.

Program - November 13, 2011 at Shangri-La Hotel, Room: Sukhotai II

Session 1: Opening and Keynote Speech

08:45-09:00 Opening

09:00-10:00 Keynote Speech by Toshiyuki Masui (Keio University)

Session 2: Long Papers

10:30-11:00 Challenges in Designing Input Method Editors for Indian Languages: The Role of Word-Origin and Context by Umair Z. Ahmed, Kalika Bali, Monojit Choudhury and Sowmya VB

11:00-11:30 Discriminative Method for Japanese Kana-Kanji Input Method by Hiroyuki Tokunaga, Daisuke Okanohara and Shinsuke Mori

11:30-12:00 Efficient dictionary and language model compression for input method editors by Taku Kudo, Toshiyuki Hanaoka, Jun Mukai, Yusuke Tabata and Hiroyuki Komatsu

Session 3: Posters and Demos

14:00-15:30 Posters and Demos

Different Input Systems for Different Devices by Asad Habib, Masakazu Iwatate, Masayuki Asahara and Yuji Matsumoto

An Accessible Coded Input Method for Japanese Extensive Writing by Takeshi Okadome, Junya Nakajima, Sho Ito and Koh Kakusho

Adaptxt demo by Keypoint Technologies

Curve-flick input for Windows Phone 7 for Japanese demo by Microsoft

Universal Text Input demo by Microsoft Research

Session 4: Short and long papers

16:00-16:20 Error Correcting Romaji-kana Conversion for Japanese Language Education by Seiji Kasahara, Mamoru Komachi, Masaaki Nagata and Yuji Matsumoto

16:20-16:40 From pecher to pêcher... or pécher: Simplifying French Input by Accent Prediction by Pallavi Choudhury, Chris Quirk and Hisami Suzuki

16:40-17:00 Phrase Extraction for Japanese Predictive Input Method as Post-Processing by Yoh Okuno

17:00-17:30 Robustness Analysis of Adaptive Chinese Input Methods by Mike Tian-Jian Jiang, Cheng-Wei Lee, Chad Liu, Yung-Chun Chang and Wen-Lian Hsu

Call for Papers (Submission due: July 25, 2011)

Methods of text input have entered a new era. The number of people who have access to computers and mobile devices is skyrocketing in regions where people cannot type their native language characters directly. It has also become commonplace to input text not through a keyboard but through different modes such as voice and handwriting recognition. Even when people type with a keyboard, it is done differently from a few years ago - adaptive software keyboards, word prediction and spell correction are just a few examples of such recent changes in text input experience. The changes are now global and ubiquitous: users are no longer willing to input text without the help of new generation input methods regardless of language, device or situation.

The challenges in text input have many underlying NLP problems in common. For example, a high quality dictionary is called for, but it is far from obvious how to construct and maintain one. A dictionary also needs to be stored in some data structure, whose optimal design may depend upon the usage. Prediction and spell correction can be very annoying if they are not smart enough. For many applications, user input can be very noisy (imagine voice recognition or typing on a small screen), so the input methods must be robust against such noise. Finally, there is no standard data set or evaluation metric, which is necessary for quantitative analysis of user input experience.

The goal of this workshop is bring together the researchers and developers of text input technologies around the world, and share their innovations, research findings and issues across different applications, devices, modes and languages. We hope that the workshop will deepen our understanding of the field as a whole, and facilitate further innovation in each application area.

We welcome participation on a wide range of topics and languages covering the text input. Topics of interest include (but are not limited to):

    • Phonetic input for the languages of the world: transliteration-based approach to non- Roman script languages
    • Input for soft keyboards and devices with hardware limitations: input systems for phones, game devices, tablet PCs
    • Investigations in new input modes: text input via speech, handwriting and/or gesture
    • Predictive input technologies: auto-completion and next word prediction for IME; spell correction
    • User feedback and community-specific data: understanding the user and usage for a better input experience
    • Models: research and practice of statistical input methods
    • Evaluation: aspects of measuring user experience in text input
    • Issues in commercial input engine deployment: data compression, cloud-based IME, user interface design

Important Dates

    • UPDATED July 25, 2011 Workshop paper submission due date NOTE: this is different from the paper submission due on the IJCNLP conference website
    • UPDATED August 12, 2011 Workshop paper acceptance notification
    • UPDATED Sept 2, 2011 Camera-ready deadline
    • Nov 13, 2011 Workshop

Submission information

Paper submission to WTIM2011 will be accepted on or before July 25, 2011 (11:59pm Samoa time, UTC-11) in PDF format via the START system:

Submissions should follow the instructions at the IJCNLP workshop page ( We accept up to 8 pages for full papers plus additional 2 pages for references. In addition to full length papers, we also encourage short papers to promote submission on a wide range of topics and languages. Short papers are up to 4 pages of content and 2 pages of references. We encourage demos at the workshop that accompany short paper submissions, but this is not a requirement. Reviewing process is blind -- do not put information that reveals the author identity in the submission.

Organizers (alphabetical order)

    • Hideto Kazawa (Google, Japan)
    • Hisami Suzuki (Microsoft Research, USA)
    • Taku Kudo (Google, Japan)

Program Committee (alphabetical order)

    • Achraf Chalabi (Cairo Microsoft Innovation Center, Egypt)
    • Frank Yung-Fong Tang (Google, USA)
    • Haifeng Wang (Baidu, China)
    • Hiroyuki Tokunaga (Preferred Infrastructure Inc, Japan)
    • Jianfeng Gao (Microsoft Research, USA)
    • Kalika Bali (Microsoft Research India)
    • Kazuma Takaoka (JustSystems, Japan)
    • Kumiko Tanaka-Ishii (University of Tokyo, Japan)
    • Mamoru Komachi (Nara Institute of Science and Technology, Japan)
    • Mike Schuster (Google, USA)
    • Monojit Choudhury (Microsoft Research India)
    • Shinsuke Mori (Kyoto University, Japan)
    • Thanaruk Theeramunkong (SIIT, Thailand)
    • Virach Sornlertlamvanich (NECTEC, Thailand)
    • Yoh Okuno (Yahoo! Japan)

Co-sponsor (Keynote speech)

    • Nara Institute of Science and Technology, Japan

E-mail contact

wtim-organizers <at>