Fast computation of Katz index for efficient processing of link prediction queries

Network proximity computations are among the most common operations in various data mining applications, including link prediction and collaborative filtering. A measure of network is Katz index, which has been shown to be best-performing path-based algorithms. With emergence very large databases, such become an important part query processing these databases. Consequently, significant effort devoted developing algorithms for efficient computation index between a given pair nodes or node every other network. Here, we present LRC-Katz, algorithm based on indexing low rank correction accelerate queries. Using variety real-world networks, show that LRC-Katzoutperforms fastest existing method, Conjugate Gradient, wide range parameter values. Taking advantage acceleration propose new exploits locality networks encountered practical applications. Our experiments resulting drastically outperforms state-of-the-art methods vanilla truncated Katz.