Semantic Knowledge in Word Completion
By Li, Jianhua; Hirst, Graeme; ASSETS 2005 - The Seventh International ACM SIGACCESS Conference on Computers and Accessibility, pp. 121-128Publication Date: October 9-12, 2005
Paper proposes an integrated approach to interactive word-completion for computer-software users with linguistic disabilities. Word completion is the task of guessing, as accurately as possible, the word that a user is in the process of typing. The model combines semantic knowledge with n-gram probabilities to predict semantically more appropriate words than n-gram methods alone. First, semantic relatives are found for English words (in this model nouns) forming the semantic knowledge base. The selection process for these semantically related words is first to rank the pointwise mutual information of co-occurring words in a large corpus and then to identify the semantic relatedness of these words by a Lesk approach, which involves looking for overlap between the words in given definitions with words from the text surrounding the word to be disambiguated. Then, the semantic knowledge is used to measure the semantic association of completion candidates with the context. Those that are semantically appropriate to the context are promoted to the top position in prediction lists. The model was evaluated with a simulated user, showing a 14.63% improvement in keystroke saving for the completion of nouns.
Published by: Association for Computing Machinery (Website:http://www.acm.org)
SIGACCESS (ACM Special Interest Group on Accessible Computing) (Web Site: http://www.sigaccess.org )
