190 likes | 298 Vues
This paper explores the concept of object-centered sociality within heterogeneous social networks, emphasizing how individuals connect through shared objects—both online and offline. Through the application of social network models and semantic web technologies such as RDF and FOAF, we analyze datasets to uncover hidden links among individuals based on common connections to objects. Results indicate that including objects in the analysis enhances our understanding of social ties. Future work aims to identify which object types facilitate these connections and their significance.
E N D
Network Analysis of Semantic Connections in Heterogeneous Social Spaces Sheila Kinsella, Andreas Harth, John G. Breslin
Overview • Object-Centred Sociality • Social Network Models • Semantic Web • Datasets • Results • Conclusions
Object-Centred Sociality • People don’t just connect, they connect through shared objects • Real life: jobs, events, activities • Online: blogs, images, web links • Objects are an important part of online sociality
Social Network Models Affiliation network One-mode network Semantic network
Semantic Web • Sir Tim Berners-Lee et al., Scientific American, 2001: • “An extension of the current web in which information is given well-defined meaning, better enabling computers and people to work in cooperation.” • Enables expression of data in a way that computers can understand • Interoperabilityand increased connectivity is possible through common formats
Resource Description Language (RDF) • Graph model used to express Semantic Web data • Each node is an instance of a class • Each link is an instance of a property (e.g. relationship) • Classes and properties defined in schemas • Namespaces indicate the schema to which classes and properties belong
Friend-of-a-Friend (FOAF) • RDF schema for describing people and the relationships that exist between them • Integrated with many other vocabularies on the Semantic Web • Can be used to express explicit and implicit links
Datasets • We start with FOAF file of Tim Berners-Lee • Extract[1] all documents within 5 links of root node • Idenfity two subgraphs of interest: • People-only network • Composed only of people directly connected to the root node, or indirectly via other people • Object-centred network • Contains all nodes who are directly or indirectly connected to the root node • Repeat for FOAF file of Andreas Harth [1] Using Semantic Web Search Engine (http://swse.org)
Effect of object inclusion: Tim Berners-Lee network Objects 361 People 691 Links 1446 People 349 Links 450 People-only network Object-centred network
Effect of object inclusion: Andreas Harth network Objects 7842 Links 47286 People 9011 People 16671 Links 24875 Object-centred network People-only network
Object and link types: Tim Berners-Lee network Most common classes Most common relationships
Object and link types: Tim Berners-Lee network Most common classes Most common relationships
Object and link types: Tim Berners-Lee network Most common classes Most common relationships
Object and link types: Tim Berners-Lee network Most common classes Most common relationships
Object and link types: Andreas Harth network Most common classes Most common relationships
Object and link types: Andreas Harth network Most common classes Most common relationships
Object and link types: Andreas Harth network Most common classes Most common relationships
Conclusions • We investigate the online social networks formed by direct interpersonal links and indirect links via objects • Dataset is from a range of semantically-enabled sources • Including objects in the network gives us additional information and reveals hidden links • Future work: • Look at which types of objects are responsible for hidden links and evaluate their relevance • Analyse object-centred network to locate sets of objects relevant to a given person