Name of the Technology : Anumaan – An open source predictive text entry system
Disability Sector : Locomotor Disability
Function : Mobility/ Holding aids, Aids for computer/ mobile or any other gadget operation
Category of research : Upgradation of existing technology/product, New technology/product (not existing in the country)
Brief description of the product/ technology:
Anumaan is a perspectivebased, onscreen predictive text entry system for GNOME desktop (Linux platform) developed at Centre for Development of Advanced Computing (CDAC), Mumbai. Anumaan is an extension of Input Method (IM) mechanism (on GNOME desktop) in the sense that it extends the notion of Input Method by including the power of text prediction.
Anumaan is designed for a (text) sequence prediction based on a certain context. The context can either be a partial word or a partial sentence. Through prediction of a certain (text) sequence which can be added to a user’s writing activity, this system helps enhance the rate of text entry. Anumaan uses N-gram language modelling technique to generate predictions. It allows the user to train it and personalize his/her writing style into different named domains. Domains trained by the user are stored as different language models to be used during the prediction process. As a user enters text (one or more characters) in the system, the context is captured and sent to the system. Then, system generates the set of predictions relevant to the context from a given domain (language model). User can then select the desired word from the set of predictions for text entry with the help of a pointing device like mouse or keyboard.
Anumaan is mainly intended to help persons with motor disabilities, specially ones facing problems in hand and finger movement. Such persons face difficulty in using regular input devices like keyboard for text entry related tasks. Anumaan can help such persons immensely in their text entry related tasks by way of predictions. It can help them augment the rate of their text entry. However, It is equally useful for common users also and can support them in creating mails, letters, documents etc. in less time.
Anumaan is developed on Java platform and supports UTF-8, so it can effectively be run on any platform. It can be easily customized for different kinds of Input devices (e.g. keyboard, mouse, joystick or any other pointing device especially built for people suffering from motor disability). In other words, it can be customized very easily to cater to different hand movement capabilities. Domains trained by the user are stored as different language models in XML format. Till now, two versions of Anumaan have been released.
• An open source predictive text entry system.
• Helpful for people with motor disability and specific learning disabilities.
• Prediction based on multiple perspectives such as domain, language features and user's writing style.
• A single comprehensive prediction list combining predictions from all relevant perspectives for easy selection.
• User friendly and simplified interface for easy interaction and navigation. Interface is designed in such a way that it reduces the visual and cognitive overhead involved in prediction selection.
• Keyboard support for selecting predictions from prediction list. It reduces time taken in selecting and committing predictions to text.
• Integrated context sensitive help.
• Support for Indian Languages.
Anumaan is available as a Free and Open Source product released under GPLv3. It can be downloaded from http://accessibledesktop.in/?q=anumaan or http://anumaan.sourceforge.net.
• There have been almost 800 downloads so far of Anumaan since its first release.
• Anumaan has also been demonstrated and deployed in over 20 disability related organisations.
Readiness level of the product/technology: Pre-productization form (beta level)
Anticipated time to reach the next readiness level: Not Available
Anticipated time for the technology/product to come in the market: 0 to 6 months
Any Hindrance(s) being faced w.r.t. the development of product/technology: Lack of user awareness
Anticipated cost: Free and Open Source Software
Dr. M. Sasi Kumar Centre for Development of Advanced Computing (CDAC)
Near Bharati Vidyapeeth,