Statistical Acquisition of Terminology Dictionary

Terminologies are specialized words and compound words used in a particular domain, such as computer science. Since they are very common in scientific articles, the ability to automatic identification of terminology could greatly assist any domain related natural language processing applications. Unfortunately, the collection of terminology information is very difficult and requires much tedious and time consuming manual work. In this paper, a semi-automatic approach is developed to extract technical words and phrases from on-line corpora. This approach can significantly reduce the manual effort in the generation of terminology dictionary. First, those domain specific words which have no entries in the universal dictionary are identified. Second, terminology words are extracted from these new words as well as the universal dictionary. Then compound words are extracted from the combination of terminology words and other words. The final computer terminology dictionary contains 1,034 words and 3,471 compound words. Experiment shows that 89.5 percent of all the occurrences of computer terminology can be identified with this terminology dictionary. keyword: Chi-square Test, Automatic Indexing, Mutual Information