Le An Ha’s PhD thesis: Advances in Automatic Terminology Processing: Methodology and application in focus

Le An Ha (2007) Advances in Automatic Terminology Processing: Methodology and application in focus. PhD Thesis, University of Wolverhampton, UK


The information and knowledge era, in which we are living, creates challenges in many fields, and terminology is not an exception. The challenges include an exponential growth in the number of specialised documents that are available, in which terms are presented, and the number of newly introduced concepts and terms, which are already beyond our (manual) capacity. A promising solution to this ‘information overload’ would be to employ automatic or semi-automatic procedures to enable individuals and/or small groups to efficiently build high quality terminologies from their own resources which closely reflect their individual objectives and viewpoints. Automatic terminology processing (ATP) techniques have already proved to be quite reliable, and can save human time in terminology processing. However, they are not without weaknesses, one of which is that these techniques often consider terms to be independent lexical units satisfying some criteria, when terms are, in fact, integral parts of a coherent system (a terminology). This observation is supported by the discussion of the notion of terms and terminology and the review of existing approaches in ATP presented in this thesis.

In order to overcome the aforementioned weakness, we propose a novel methodology in ATP which is able to extract a terminology as a whole. The proposed methodology is based on knowledge patterns automatically extracted from glossaries, which we considered to be valuable, but overlooked resources. These automatically identified knowledge patterns are used to extract terms, their relations and descriptions from corpora. The extracted information can facilitate the construction of a terminology as a coherent system. The study also aims to discuss applications of ATP, and describes an experiment in which ATP is integrated into a new NLP application: multiple- choice test item generation. The successful integration of the system shows that ATP is a viable technology, and should be exploited more by other NLP applications.


   author =   {Le An Ha},
   title =    {Advances in Automatic Terminology Processing: Methodology 
               and application in focus},
   school =   {School of Humanities, Languages and Social Sciences, 
               University of Wolverhampton},
   year =     {2007},
   address =  {Wolverhampton, UK},
   month =    {July},
   URL =      {http://clg.wlv.ac.uk/papers/ha-thesis.pdf}