Heterogeneous Identity Expression and Association Method Based on Attribute Aggregation

Existing identity expression methods are often limited in a single security domain, and this is inadequate to meet the cross-domain access requirements of heterogeneous networks. In view of this problem, we propose an index system for the ubiquitous expression of heterogeneous identities, and introduce the concept pair matching based attribute aggregation method by combining the characteristics of heterogeneous identity alliances. The selection of concept pairs considers the original meaning of attribute characteristics, including the lexical level, i.e., class, ontology, label, description, the structural level, i.e., position, distance between nodes, and the semantic level, i.e., formal concept analysis. As for the attribute aggregation, if multiple attributes from a heterogeneous network contain the same or similar concepts, they are considered the same attribute for the user identity in a heterogeneous network. Relevant domain knowledge or heuristic knowledge will adjust the result of attribute aggregation, and the constraint relationship between conceptual structures are used to adjust and optimize the attribute aggregation set. Based on the identity attribute index system of the heterogeneous identity alliance, the identity similarity evaluation results based on each attribute are generated. When the comprehensively considered identity similarity evaluation result is higher than the empirical threshold, the heterogeneous identity alliance has different trusts for the same user. The experimental results show that our scheme has a better overall aggregation effect on identity attribute aggregation.