Evaluating the role of community detection in improving influence maximization heuristics

Abstract

Both community detection and influence maximization are well-researched fields of network science. Here, we investigate how several popular community detection algorithms can be used as part of a heuristic approach to influence maximization. The heuristic is based on the community value, a node-based metric defined on the outputs of overlapping community detection algorithms. This metric is used to select nodes as high influence candidates for expanding the set of influential nodes. Our aim in this paper is twofold. First, we evaluate the performance of eight frequently used overlapping community detection algorithms on this specific task to show how much improvement can be gained compared to the originally proposed method of Kempe et al. Second, selecting the community detection algorithm(s) with the best performance, we propose a variant of the influence maximization heuristic with significantly reduced runtime, at the cost of slightly reduced quality of the output. We use both artificial benchmarks and real-life networks to evaluate the performance of our approach.

Privacy-Aware and Secure Decentralized Air Quality Monitoring

“…Regarding the methodology used, it is important to note that, due to the submodularity of the problem we introduced, the method in the area of infection models performs within a constant factor of 1 − 1/e as good as the optimal solution. The basic problem was further investigated in our latest research, which is, however, a generalized solution that can be used in a decentralized solution [82]. The main problem of the greedy algorithm is that it provides an insufficient running time in the case of big networks, and it is not efficiently scalable.…”

Network based data oriented methods for application driven problems

“…In the third chapter, we further examined the infection processes and presented a methodology related to the topic of community detection and infection maximization [76,77]. In one hand, the method is intended to increase the efficiency of the original greedy algorithm by reducing the search space, in the other hand the obtained results provide a methodology to compare and benchmark overlapping community detection algorithms from infection point of view.…”

“…A harmadik fejezetben a fertőzési folyamatokat vizsgáltuk tovább, valamint egy a közösségkereső és fertőzésmaximalizáló algoritmusok témaköréhez kapcsolódó módszertan került bemutatásra [76,77]. Egyrészről a módszer a mohó algoritmus hatékonyságát hivatott növelni a keresési terének csökkentésével, másrészről az így kapott eredmények lehetőséget adnak átfedő közösségkereső algoritmusok fertőzésterjedési szempontból történő összehasonlítására illetve benchmarking módszertan felállítására.…”

“…After the concept is introduced, we test the methods from both sides on real and randomly generated networks, and give comparison on existing community detection algorithms. The methodology is studied in [76,77] Chapter 4, introduces an infection maximization model based on the Generalized Independent Cascade [29], using a use case from psychology. In the first part of this chapter, we focus on the new infection maximization model and a method to minimize the network effect in a case of an existing influence or infection.…”