论文部分内容阅读
【目的】借鉴Lesk词义消歧思想,提出并实现一种利用术语定义来发现汉语同义词的方法。【方法】将新能源汽车领域汉语科技词系统中的术语及其定义作为测试集,首先对术语定义做分词和词性标注,并进行人工校对,然后抽取出动词和名词词性的实词,再根据两个术语定义中相同的实词数量及位置信息计算术语的相似度,最后根据相似度和给定的阈值得到同义词关系的推荐。【结果】利用准确率、召回率、F值对同义词发现效果进行评价,论证该方法的有效性,结果表明该方法可以达到较高的准确率,但是召回率比较低。【局限】该同义词发现方法不能剔除反义关系和相关关系的术语对,造成召回率较低。【结论】该方法较为简便快捷有效,并且可达到较高准确率,但召回率有待提高。
[Objective] To draw lessons from Lesk meaning disambiguation and to propose and implement a method to discover Chinese synonyms by using the definition of terms. 【Method】 Using the terms and definitions of Chinese science and technology word system in the field of new energy vehicles as a test set, word definitions and part of speech tagging are first defined and manually calibrated. Then the real words of the verbs and noun parts are extracted, The same number of terms in the definition of terms and location information to calculate the similarity of terms, and finally get the recommendation of the synonym relationship according to the similarity and the given threshold. 【Result】 The results of synonym discovery were evaluated by accuracy, recall and F value. The results show that this method can achieve high accuracy, but the recall rate is relatively low. [Limitations] This synonym discovery method can not eliminate term pairs of antisense and related relationships, resulting in lower recall rates. 【Conclusion】 The method is simple, quick and effective, and can achieve high accuracy, but the recall rate needs to be improved.