A Trusted Behaviour Learning for Interest Prediction inSocial Ontology Based on Weighted Graph

J. Kalavathi, S.Balamuraliand M. Venkatesulu


Objective - In this paper, an attempt is made to suggest a social ontology dependent behavior learning approach to anticipate the interest of group of users utilizing social graphs and density measure.

Methodology/Technique - Weighted clustering is implemented using trustness to ensure better linking among the social domain. The social ontology includes a set of classes in which related terms and properties are referred regarding a category or interest. The suggested approach clusters the user groups into N number of weighted clusters of Ontology set S, and there would be a weight cluster for each one of the interests or subjects or classes defined in the social ontology. From the social data set, a social graph can be built where every user may be regarded as a node and there would be a connection between user solely if they have general interest. The suggested technique clusters the user groups based on the trust value which is mostly dependent on density measures.

Findings - The evaluation results have shown that the proposed technique has produced less time and space complexity values.

Novelty - The proposed technique identifies the user group interest in exact way and acquires effective results

Type of Paper: Review

Keywords: Social Ontology; User Interest; Density Measure; Behavior Learning; Weighted Clustering; Trusted Graphs

DOI: https://doi.org/10.35609/gjetr.2016.1.1(5)

Global Journal of Engineering & Technology Review Conferences 2019-20:
Meet Inspiring Speakers and Experts at our 1000+ Global Annual Meetings