Optimum Session Interval based on Particle Swarm Optimization for Generating Personalized Ontology S. Chitra 1* and B. Kalpana 2 1 Department of Computer Science, Government Arts College, Coimbatore, Tamil Nadu, India; schitra789@gmail.com 2 Avinashilingam Institute for Home Science and Higher Education for Women, Coimbatore, Tamil Nadu, India Abstract A semantic web usage mining method is suggested to identify the association between consumer emotions and buying behaviors by utilizing the web log data. Fuzzy logic is used to signify the temporal conception and resource attributes for the requested URLS of web access activities. From this, a Personal Web usage Ontology is created which facilitates semantic web applications. But the limitation is less efficient in terms of accuracy and user satisfaction. Thus an innova- tive technique which is called Optimum Session Interval based Particle Swarm Optimization (OSIPSO) is introduced. This technique is used to find the optimum session interval. The Particle swarm optimization has no overlapping and mutation computation and it is proficient in global search. Additionally, an associative classification is used to enhance the accuracy. Associative classification is a combination of associative rule mining and classification rule mining. An experimental result shows that the proposed work has a high accuracy and high efficient. Keywords: Associative Classification, Emotion and Behavior Profiling, Ontology Generation, Semantic Web, Particle Swarm Optimization *Author for correspondence 1. Introduction Web usage mining is an automatic detection of patterns in click streams and related data are collected as a result of user relations with one or more Web sites. he main intent of web usage mining is to observe the behavioral patterns of users interrelating with a web site. he discov- ered patterns are generally characterized as a collection of pages, objects or resources which are regularly accessed by groups of users with common interests. Human motions are a signiicant factor of human behaviors in web mining analysis 1 . he relations between customer sensations and their buying behaviors have been well recognized 2,3 . In the web applications, the consumer emotions and behav- iors are signiicant to enhance the performance. he customers self-report is used to analyze the emo- tions of the particular resources. Ater the user visit the website there exists an opportunity for the users to record their suggestions in the emotional state. By utilizing the records provided by the users it is possible to analyze the emotional inluence of the users. Web usage mining is an essential approach which is used to detain the consumer access pattern and to realize recurrent user access patterns 4 . A semantic web usage mining is a method whereas it is used to relate each demanded webpage with one or more ontologies for better understanding of web navigation. A semantic web usage mining 5 is one of the techniques which is used to automatic creation of periodic web access pattern. Earlier, web usage mining methods 6 are focused on mining frequent access patterns which have occurred recurrently within the entire duration of all the user access sessions. So, this method analyzes the frequently used resources at a particular time period. Furthermore, ontology is generated to collect web access behaviors and emotional inluence of the users for the speciic resources. To improve the classiication accuracy in the proposed work, Optimum Session Interval based on Particle Swarm Optimization (OSIPSO) is introduced to identify the Indian Journal of Science and Technology, Vol 7(8), 1137–1143, August 2014 ISSN (Print) : 0974-6846 ISSN (Online) : 0974-5645