期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：12
页码：2698
出版社：Journal of Theoretical and Applied
摘要：In the apparent, biomedical literature will be retrieved starting with the mixed literature articles, utilizing Multi-Kernel Fuzzy c Means (MK-FCM) clustering technique. Literature retrieval or document retrieval is an action that utilization professional techniques for medically examine papers retrieval, the report card also different information to move forward research and practice. In the work, the biomedical query article is made also preprocessing system may be utilized and there would two sorts of preprocessing to be specific stop words removal and stemming. Then afterward those executions for preprocessing the retrieval techniques, for example, vector space modeling and retrieval modeling is utilized. In the vector space modeling the term frequency and inverse frequency measure may be broken down after that on retrieval modeling the SMART and BM25 standard models used to furthermore retrieve the literature. Behind this Multi-Kernel Fuzzy c Means (MK-FCM) technique is utilized for clustering the literature. In this differentiate kernel function named as FCM, Linear FCM, Quadratic FCM and Composite kernel function will be investigated to several measuring tests. Those biomedical query article databases where brain tumor, breast cancer, kidney stone and neovascularization is clustered. For single, randomly two, randomly three and overall four types of input biomedical documents the precision of the composite kernel function is 90% and recall of composite kernel function is 87% were compared with different strategies and it progresses the level best. By using these techniques, the retrieval performed better from other techniques.
关键词：Biomedical literature; Multi-Kernel Fuzzy c Means (MK-FCM) clustering; Composite kernel function; Linear FCM and Quadratic FCM.
