基于引文耦合分析方法的相关词识别

Relevance Terms Recognition Based on Bibliographic Coupling Analysis Method

借鉴引文耦合分析方法,将词条定义中的实词比作词条的参考文献,根据词条定义中实词耦合强度实现相关词的识别。首先对词条定义进行分词和词性标注,并进行人工校对,然后抽取出动词和名词词性的实词,以词条定义中实词的耦合强度作为判定标准实现相关词的推荐,并用人工校对的方法,计算相关词识别的准确率、召回率和F值,论证该方法的有效性。该实验将新能源汽车领域汉语科技词系统中随机选择的500条词条及其定义作为测试集,发现该方法可以达到较高的准确率和召回率。

Enlightened by citation coupling analysis method, regarding the content words in the definition of term as the term＇s refer-ences, according to the content words coupling strength of the term＇s definition, the relevance terms recognition is achieved. First, the Chinese word segmentation, part-of-speech tagging and manual correction of term definition are processed. Then, verbs and nouns con-tent words are extracted and content words coupling strength is regarded as the criterion to achieve the relevance terms recognition. At last, manual correction is used to calculate the precision and recall of relevance terms recognition to demonstrate the effectiveness of this meth-od. This experiment regards the Chinese scientific and technical vocabulary system＇s 500 randomly selected terms and their definitions as the test set （ in the field of new energy vehicles） and find that the method can achieve a high precision and recall.