Towards The Pattern Of Definition In English Learners’ Dictionaries Based On Categorization

Definition has always been considered as the core of a dictionary, buttraditionally, the entries in the dictionaries are arranged in the alphabetical order,which leads to the fragmentation of semantic network among the lexical items. Thus,words related by various kinds of lexical relations are separated and get differentdefinition patterns, which may cause confusion for the dictionary users and lowertheir cognitive efficiency.Categorization is a basic mental process of human beings to understand theworld and it is fundamental in language, since people group objects and ideas in thereal world into different categories and then form concepts in the mind, during whichlanguage plays a central role. Therefore, we adopt the categorization approach toimprove the definitions.In the present study, we choose the3000core vocabulary as the research subjects,and we discuss the construction of a database based on lexical relations in order tocategorize the3000core vocabulary, including synonymy, antonymy, meronymy,hyponymy, troponymy, frequent co-occurrence, framework, etc. Besides, weintroduce some lexical data resources on the Internet that help us analyze the lexicalrelations, such as the WordNet, FrameNet, Dante and Sketch Engine. Based on thelexical relation database, we choose the synonymy categories of nouns, verbs andadjectives and analyze their definitions in the present dictionaries. With the help ofthese definitions and the lexical data resources, we manage to figure out the prototypeof each category and the non-prototypical features of each member of the category.Finally, we attempt to bring forward a new definition pattern which illustrates theprototypical feature of the category as well as the distinctive non-prototypical featuresof each member.Through the optimization of the definition pattern, we hope that we can improvethe dictionary users’ cognitive efficiency and at the same time save the precious space that is meant for the “synonym discriminate” part. Furthermore, the lexical relationdatabase based on categorization has a practical value and a promising prospect inelectronic dictionaries. In the present study there exist certain limitations, and wehope the further study can make some progress and improvement.