HNRWalker: recommending academic collaborators with dynamic transition probabilities in heterogeneous networks

摘要: Multi-source information not only helps to solve the problem of sparse data but also improves recommendation performance in terms personalization and accuracy. However, how utilize it for facilitating academic collaboration effectively has been little studied previous studies. Traditional mechanisms such as random walk algorithms are often assumed be static which ignores crucial features linkages among various nodes multi-source networks. Therefore, this paper builds a heterogeneous network constructed by institution co-author proposes novel model collaborator recommendation. Specifically, four neighbor relationships corresponding similarity assessment measures identified according characteristics different network. Further, an improved algorithm known “Heterogeneous Network-based Random Walk” (HNRWalker) with dynamic transition probability new rule selecting candidates proposed. According our validation results, proposed method performs better than benchmarks improving performances.