News & Events
|
Topic: Research Seminar [3:30pm 05/06/2006]: " Automated semantic assistance for human translators" and "Extraction and Compositionality Measurement of Multiword Expression"
|
 Posted - 17/05/2006 : 15:46:02
City University of Hong Kong
Department of Chinese, Translation and Linguistics
Research Seminar
Automated Semantic Assistance for Human Translators
Presented by
Dr. Paul Rayson
Lancaster University (UK)
Abstract
The ASSIST project aims to address the problem of providing contextual examples of translation equivalents for words from the general lexicon. The ASSIST tool will suggest contextual examples in the target language that are semantically and syntactically related to a selected example in the source language. The project is exploiting comparable corpora and developing semantic annotation that is uniform for the source and target languages. In our case, we are focussing on the English-Russian language pair. In this talk, I will describe the ASSIST framework and the construction of the semantic taggers in each language.
The ASSIST project is supported by two UK-EPSRC grants: EP/C004574 for UCREL, Lancaster University, and EP/C005902 for Centre for Translation Studies, University of Leeds. For more details, see: http://ucrel.lancs.ac.uk/projects/assist/.
Speaker
Dr Paul Rayson is director of the UCREL (University Centre for Computer Corpus Research on Language) research centre at Lancaster University (UK). His research interests are in the area of corpus linguistics (and related subjects within natural language processing and language engineering) and the application of corpus-based methods in word frequency dictionaries, semantic analysis, information extraction, systems engineering and decision management.
Extraction and Compositionality
Measurement of Multiword Expressions
Presented by
Dr. Scott Songlin Piao
Lancaster University (UK)
Abstract
Recently the issue of extraction and processing multiword expressions (MWEs) is receiving an increasing attention, evidenced by a series of MWE workshops. It is a tough challenge for both corpus linguistics and NLP to automatically identify, classify and translate MWEs. For example, in the corpus-based MT study, efficient methods and algorithms have yet to be developed for detecting MWEs in a given context and map them to correct translation equivalent(s). In this talk, I will present our recent work on automatic detection and compositionality measurement of MWEs, in which both statistical and lexical approaches have been tested. Based on the results of our experiment, we claim that 1) statistical algorithms and linguistic approaches can provide complementary methods for MWE extraction, and 2) semantic lexical resources provide another approach for measuring compositionality in addition to the statistical ones.
Speaker
Dr Scott Songlin Piao is research fellow at UCREL. His research interests include parallel text alignment, corpus processing and developing algorithms for language engineering tasks. He programs in C, C++ and Java.
Date: 05 June 2006, Monday
Time: 3:30pm - 5:30pm
Venue: B7603 (Lift 3, 7/F, Blue Zone), Academic Building, CityU of HK
~All Are Welcome ~
|
|
|
|
Enquiry: LTenquiry@cityu.edu.hk
