Abstract—Social networks have recently gained a growing interest on the web. Traditional formalisms for representing social networks are static and suffer from the lack of semantics. In this paper, we will show how semantic web technologies can be used to model social data. The SemTemp ontology aligns and extends existing ontologies such as FOAF, SIOC, SKOS and OWL-Time to provide a temporal and semantically rich description of social data. We also present a modeling scenario to illustrate how our ontology can be used to model social networks. Keywords—Ontology, semantic web, social network, temporal modeling. I. INTRODUCTION OCIAL NETWORKS (SN) have received a surge of attention in the recent years. This can be explained by the fact that these social networks allow users to connect to each other and to share and exchange diverse kinds of information; Thiers posts, activities, events and interests among them [1]. The exponential growth of Social Networks makes it more interesting and so hard to analyze and to take use of it, because the large amount of social data is unstructured. That recognizes the need for new models for representing social network. Traditional formalisms generally use graphs to represent the social structure [2]-[4]; the nodes models social actors while links models relations between them. These formalisms suffer from a range of problems. They luck from semantics and a lot of information about nodes and links are ignored. They are less expressive and represent only simple networks and do not take into account heterogeneity of nodes and links. In a real social network, like Facebook for example, the nodes can represent individuals, organizations, resources, etc., and the links can represent various types of relationship (e.g. friendship, family, colleague, etc.). The individuals have also different roles and different status. Another issue concerns the dynamic aspect. The social structure evolves over time. Individuals can join or leave the network. Relationships can change also; they can be added or removed from the network. The temporal evolution of the social network is very important and can provide enhancements in social network analysis. However, most proposals provide a static description and Souâad Boudebza is with the University of Mohamed Seddik Ben Yahia – Jijel, Department of Computer Science, Jijel, Algeria (e-mail: s_boudebza@esi.dz). Omar Nouali is with the Department of Research Computing, CERIST, Algiers, Algeria (e-mail: onouali@cerist.dz). Fiaçal Azouaou is with the Ecole Nationale Supérieure D'informatique, ESI, Algiers, Algeria (e-mail: f_azouaou@esi.dz). dynamic of social network is frequently overlooked. Other serious problem relies to interoperability. Indeed, the existing models are not suitable for exchanging data between multiple social applications. One of the primary goals of Semantic Web is to promote integration and interoperability [5]. Ontology form a vital component in the semantic web by building a formal representation that can provide meaningful description and linkage across data. We aim in this work to use semantic web technologies to provide a temporal semantic representation of social networks. The main contribution is the development of SemTemp ontology that extends and aligns existing vocabularies [6]-[8], [12]. This contribution is detailed as follows: Section II deals with the related work done in social networks modeling. Section III details the development the SemTemp ontology. Section IV describes the implementation of our ontology. Section V shows some modeling examples that illustrate usage of the ontology. Section VI, concludes the paper. II. RELATED WORK The first representation of social networks has been proposed in the early 1930s by Moreno [9] and called “sociogram”. It provides a graphical representation of the social structure; the individuals are represented by circles, rectangles and relationships by liens. It represents only relationships of attraction or repulsion. This representation is adopted for restricted groups and becomes unreadable for wide networks. In the middle of the twentieth century, graph theory [2] has become the conventional representation of social networks. A graph consists of a set of points and lines (oriented or non- oriented, weighted or unweighted, labeled or unlabeled) connecting tow points, called respectively vertices and edges. Several graph based-models are proposed. The Oriented graphs are adopted for representing social networks with symmetric relationships, like “friend” and “family” relationships in Facebook. In contract, nonsymetric relationships, like “Follow” in Twitter are modeled using non oriented graphs. Weighted graphs are often used to model networks where links have different intensity levels. The weights on edges denote the occurrence of interactions (e.g. number of messages, or comments) between people. Labeled graphs are well suited to model social networks with different types of relationships. In Facebook for example, the labels: friend, family, favorite, etc. are used to type relationships. Bipartite graphs are commonly used to model networks using two types of nodes, like content-sharing sites Flickers, Ontology-Based Approach for Temporal Semantic Modeling of Social Networks Souâad Boudebza, Omar Nouali, Faiçal Azouaou S World Academy of Science, Engineering and Technology International Journal of Computer and Information Engineering Vol:10, No:1, 2016 176 International Scholarly and Scientific Research & Innovation 10(1) 2016 ISNI:0000000091950263 Open Science Index, Computer and Information Engineering Vol:10, No:1, 2016 publications.waset.org/10003513/pdf