	LOGIN
	>> Home
	>> Topics
	>> Students
	>> Partners
	>> Statistics

Information for topics

Topic Id:

ID topic:	405
Partner Email:	ktutuncu@selcuk.edu.tr
Project Title:	SEARCHING FOR SIMILAR DOCUMENTS USING FUZZY CLUSTERING
Abstract:	In this study, term weighting methods will be emphasized related to preprocessing stage. Because of using fuzzy clustering, the usage of existing term weighting methods with fuzzy clustering will be investigated and their performances will be compared.The method which shows the best performance will be determined and this method will be used in the following stages. For similarity measurement stage, the performances of existing similarity measurements in suggested search approach, will be investigated. Still for this stage, a new similarity measurement that bases on the size of data will be suggested. As last, multiple category problems that can be summarized as a test document belonging to more that one category, will be discussed. Clustering/classification stage of the suggested search approach for solution of this problem will be tried to develop. For this aim, existing classification method will be adapted to the problem to determine which documents belong to more than one category. Besides, the categories will be tried to determine by being formed a relation matrix, if a document belongs to more than one category.
Advisor:	Prof. Dr. Novruz ALLAHVERDI
Link:
Degree:	Master
Keywords:
Computer Software
Algorithms & problem solving
Artificial intelligence & Neural networks
Data mining
Data types & Data structures
Information systems