Finding Heavy Paths in Graphs: A Rank Join Approach (1112.1117v2)

Abstract: Graphs have been commonly used to model many applications. A natural problem which abstracts applications such as itinerary planning, playlist recommendation, and flow analysis in information networks is that of finding the heaviest path(s) in a graph. More precisely, we can model these applications as a graph with non-negative edge weights, along with a monotone function such as sum, which aggregates edge weights into a path weight, capturing some notion of quality. We are then interested in finding the top-k heaviest simple paths, i.e., the $k$ simple (cycle-free) paths with the greatest weight, whose length equals a given parameter $\ell$. We call this the \emph{Heavy Path Problem} (HPP). It is easy to show that the problem is NP-Hard. In this work, we develop a practical approach to solve the Heavy Path problem by leveraging a strong connection with the well-known Rank Join paradigm. We first present an algorithm by adapting the Rank Join algorithm. We identify its limitations and develop a new exact algorithm called HeavyPath and a scalable heuristic algorithm. We conduct a comprehensive set of experiments on three real data sets and show that HeavyPath outperforms the baseline algorithms significantly, with respect to both $\ell$ and $k$. Further, our heuristic algorithm scales to longer lengths, finding paths that are empirically within 50% of the optimum solution or better under various settings, and takes only a fraction of the running time compared to the exact algorithm.