Eng · 繁體 · 简体

News & Events

 News & Events Home
 News and Events Archive
Research Degree Forum: "Automatic Evaluation of Machine Translation via Word Choice and Word Order"
Jump To:
 
Topic:  Research Degree Forum: "Automatic Evaluation of Machine Translation via Word Choice and Word Order"
posted itemPosted - 08/06/2010 :  11:24:14

Department of Chinese, Translation and Linguistics

 

Research Degree Forum

 

 Automatic Evaluation of Machine Translation via Word Choice and Word Order

 Presented by

Mr. WONG Tak Ming

PhD candidate, Department of Chinese, Translation and Linguistics, City University of Hong Kong

 

Date: 29 June 2010, Thursday
Time: 4:30 - 5:30 pm
Venue:
B7603 (7/F, Blue Zone), Academic Building, CityU

 

 

Abstract

 

This talk will present our recent work on developing new automatic evaluation measures for machine translation. The current practice in the field focuses on computing the textual similarity between machine translation outputs and corresponding human translations (used as gold-standard references). But this kind of approaches has been arousing criticisms on their theoretical validity and practical performance. In our approach we resort to the most fundamental factors of a sentence, namely, the word choice and word order, and provide explicit quantification for each of them. The former is assessed in terms of word forms at various linguistic levels, including surface form, stem, sound and sense, and of weighing the informativeness of each word. The latter is quantified in terms of the discordance of word position and word sequence between a translation candidate and its reference. The potential applications of our work in various related disciplines are also discussed.

 

 

 

Speaker

 

Mr. WONG Tak Ming is currently a PhD candidate from the Department of Chinese, Translation and Linguistics. His research focuses on the evaluation of machine translation, in particular, its automated metrics and the theoretical studies of the relationship between human and automatic measures. On this basis, the study on proper selection and use of machine translation technology for lay users is also explored.

 

 

 

~ CTL Staff and Research Degree Students only ~






   

 

Enquiry: LTenquiry@cityu.edu.hk