CAJ | 학술논문

基于本体的文本聚类方法,在文本表示上引入WordNet,并定义了关键概念集,使用WordNet中的概念节点及概念间的语义关系减少文本特征向量维数,提高聚类效果. 聚类过程中,算法使用文本的关键概念集和概念特征向量计算文本相似度,利用文本的关键概念集标注聚簇为聚类结果中的各个簇提供解释. 实验结果表明,该方法有效地减少了文本特征向量的维数,提高了文本聚类效果以及聚类结果的可解释性.
기우본체적문본취류방법,재문본표시상인입WordNet,병정의료관건개념집,사용WordNet중적개념절점급개념간적어의관계감소문본특정향량유수,제고취류효과. 취류과정중,산법사용문본적관건개념집화개념특정향량계산문본상사도,이용문본적관건개념집표주취족위취류결과중적각개족제공해석. 실험결과표명,해방법유효지감소료문본특정향량적유수,제고료문본취류효과이급취류결과적가해석성.
The text clustering method based on ontology applies WordNet and key concept set during text reprensentation,and the concept nodes and the semantic relations between the concepts in the ontology WordNet are used to reduce the number of features so as to improve clustering effect. And during text clustering,the algorithm uses the key concept set and the concept feature vector to calculate the similarity and uses key concept set to provide an explanation for every cluster of the result. The experimental results show that the method can effectively reduce the dimension number of the text feature vector and improve the text clustering effect compared with other text clustering algorithm and the novel method for text clustering can come up with a good explanation for the clusters.