Word Prediction in a Running Text: A Statistical Language Modeling for the Persian Language

Proceedings of the Australasian Language Technology Workshop 2005, pages 57–63, Sydney, Australia, December 2005. Word Prediction in a Running Text: A Statistical Language Modeling for the Persian Language Masood Ghayoomi Seyyed Mostafa Assi Institute for Humanities and Cultural Studies, Tehran, 14374 Institute for Humanities and Cultural Studies, Tehran, 14374 masood29@yahoo.com s_m_assi@ihcs.ac.ir Abstract Word prediction is the problem of guessing which words are likely to follow in a given segment of a text to help a user with disabilities. As the user enters each letters of the required word, the system displays a list of the most probable words that could appear in that position. In our research we designed and implemented a word predictor for the Persian language. Three standard performance metrics were used to evaluate the system including keystroke saving, the most important one. The system achieved 57.57% saving in keystrokes. 1. Introduction A word prediction system facilitates typing of a text for a user with physical or cognitive disabilities. As the user enters each letter of the required word, the system displays a list of the most likely completions of the partially typed word. As the user continues typing more letters, the system updates the suggestion list accordingly based on the new context. If the required word is in the list, the user can select it with a single keystroke. Then, the system tries to predict the next word. It displays a list of suggestions to the user. If he finds the next intended word, he selects it; otherwise he enters the first letter of the next word to restrict the suggestions. The process continues to complete the text. For someone with physical disabilities, each keystroke is an effort; as a result, the prediction system saves the user's energy by reducing his physical effort and also the system assists the user in the composition of the well-formed text qualitatively and quantitatively (Fazly, 2002). Moreover, the system increases user’s concentration (Klund and Novak, 2001). Traditionally, word predictors have been built based on statistical language modeling (SLM) (Gustavii and Pederssen, 2003). SLM is based on the probability of a sequence of n given words (n-gram). A number of word prediction systems are available today for English, Swedish, and other European languages. Most of these systems have used n-gram language modeling. The current research deals with the design and implementation of a word prediction system based on SLM for the Persian language. 2. Related Works By looking back, early prediction systems mostly were developed in the 1980s. They were used as a writing assistance system for the one with disabilities. In the early systems, they only suggested the high frequency words that matched the partially typed word and ignored the entire previous context (Swiffin et al, 1985). SoothSayer is such a system. To make suggestions more appropriate, some systems look at a larger context by exploiting word bigram language model beside the word unigram. WordQ (Nantais et al, 2001; Shein et al, 2001) is a system which is developed for English. …

Content truncated to 500 words