Optimization Of The Setting Of The Near-synonym Discrimination Boxes In ECLDs

English synonyms,particularly near-synonyms,as they are generally large in number,play a very important role in vocabulary acquisition.The near-synonym discrimination box,a crucial component of ECLDs,is a valuable resource for Chinese EFL learners to retrieve near-synonym discrimination information,however,the setting of the near-synonym discrimination boxes is far from being satisfactory.In this thesis,the issue of setting near-synonym discrimination boxes is addressed in three interrelated aspects:the range of the cluster,the multi-dimensional discrimination information and the access system.Six theories are introduced and integrated to lay the theoretical foundation for the present study.They are Semantically Synonymous Field,Prototype Theory,Seven Types of Meaning,Semantic Prosody,Prominence Principle and User Perspective.A synchronical and diachronical survey of the setting of the near-synonym discrimination boxes in typical ELDs and one ECLD reveals that in the practice of setting the near-synonym boxes,ECLDs lag behind ELDs,especially in applying corpus evidence to the setting of near-synonym discrimination boxes in ECLDs.Corpus-based approaches are adopted in determining the range of the cluster and in discriminating near-synonyms.The register,significant collocates and the semantic prosody of the near-synonyms are investigated in corpora.Eight ways are proposed to integrate the corpus-based findings into the near-synonym discrimination boxes.Based on the strong points found in ELDs and the corpus-based approaches, suggestions for improving the setting of the near-synonym discrimination boxes are made.Models for the optimized near-synonym discrimination boxes and the external access system of the near-synonyms are presented as well.