CAJ | 학술논문

针对HITS算法、传统最大流算法在挖掘Web社区时存在主题漂移、噪音页面等问题，采用基于传递概率的边容量分配最大流改进算法，开发了一个改进的Web社区结构挖掘系统，详细描述了该系统的设计和实现过程。实验表明，利用该系统进行Web社区挖掘能较好的解决传统算法中存在的问题，进一步提高了Web社区挖掘的准确性。
침대HITS산법、전통최대류산법재알굴Web사구시존재주제표이、조음혈면등문제，채용기우전체개솔적변용량분배최대류개진산법，개발료일개개진적Web사구결구알굴계통，상세묘술료해계통적설계화실현과정。실험표명，이용해계통진행Web사구알굴능교호적해결전통산법중존재적문제，진일보제고료Web사구알굴적준학성。
For the topic drift, noise pages and other problems of the HITS algorithm and the traditional maximum flow algorithm in mining Web community, maximum flow improvement algorithm based on transmission probability's side capacity assignment is used, an improvement Web community structure mining system is developed,and described this system design and the realization process in detail. It is proved with numbers of experiment that the system of the community structure mining can well solve problems of traditional algorithm,the accuracy of the Web community mining is more improved.