Impact of Consolidating Web Based Social Networks on Derived Trust Factors

Individuals are typically members of a variety of web-based social networks (both explicit and implied), but existing trust inference mechanisms typically draw only on a single network in order to calculate trust between any two individuals. This both reduces the likelihood that a trust value can be calculated (as both people have to be members of the same network), and reduces the quality of any trust inference that can be drawn (as it will be based on only a single network, typically representing a single type of relationship). To make trust calculations on multiple distributed (MuDi) social networks those networks must be first consolidated into a single network. This is challenging as simple consolidation strategies such as summing or averaging trust values tend to distort trust values. In this paper we present an analysis of different consolidation approaches for MuDi trust networks, and propose a Weighted Ordered Weighted Averaging (WOWA) approach, that avoids these problems by including the value added (by a given tie in a given network) as a factor during consolidation. We evaluate the consolidation strategies using a numerical simulation, where we generate a range of networks that simulate the characteristics of real-world networks, and analyse what happens to trust factors such as average strength of trust ties (TS ), and average length of trust path (TL), as they are consolidated with different node and tie overlap. We discover that while simple strategies such as summing or averaging distort the trust values in the consolidated network by amplifying or dampening trust metrics, the WOWA approach maintains the integrity of the trust metrics between individuals, and can dramatically increase the potential number of trust paths.