Social Alignment for Preprocessing User-Constructed Ontologies

User-constructed ontologies have caused some difficulties in being discriminated from other user-constructed ontologies. For efficiently annotating resources in their own repositories, people simply append ontology fragments retrieved from standard ontologies and from other neighbors' ontologies along to social links. In this paper, we propose a preprocessing method to extract preferential concepts for comparing with social semantics. In order to prune out irrelevant concepts from user-constructed ontologies, alignment-based concept classification process is designed by checking these two main criteria; i) redundancy (e.g., if there already exist semantically identical concepts), and ii) tendency (e.g., if there exist semantically declined concepts). Finally, we want to show an application scenario to demonstrate our contributions.