A new learning automata‐based sampling algorithm for social networks

Abstract

Recently, studying social networks plays a significant role in many applications of social network analysis, from the studying the characterization of network to that of financial applications. Due to the large data and privacy issues of social network services, there is only a limited local access to the whole network data in a reasonable amount of time. Therefore, network sampling arises to studying the characterization of real networks such as communication, technological, information, and social networks. In this paper, a sampling algorithm for complex social networks that is based on a new version of distributed learning automata (DLA) reported recently called extended DLA (eDLA) is proposed. For evaluation purpose, the eDLA-based sampling algorithm has been tested on several test networks and the obtained experimental results are compared with the results obtained for a number of well-known sampling algorithms in terms of relative error and Kolmogorov-Smirnov test. It is shown that eDLA-based sampling algorithm outperforms the existing sampling algorithms. Experimental results further show that the eDLA-based sampling algorithm in comparison with the DLA-based sampling algorithm has a 26.93% improvement for the average of Kolmogorov-Smirnov value for degree distribution taken over all test networks. Therefore, researchers try to study and analyze real networks such as communication, technological, information and OSNs via some metrics (i.e., centrality measures) or some techniques (i.e., network sampling methods) to estimate the characterization of these networks. Besides, some hidden important properties of networks (e.g., user age distribution, user activities, user connectivity, and many more in OSN) are discovered by studying on small sampled data. It is noted that sometimes sampling methods applied on the networks subject to preserve vital properties of the networks to reach a suitable visualization.A social network can be represented as a graph with a set of nodes such as users and a certain type of relationship between users as edges of graph. Let G = hV, Ei be a given social network graph, where V is the set of nodes and E is the set of edges. Sampled network G s = hV s , E s i is the induced subgraph of G by sampling rate ϕ, where V s ⊆ V, E s ⊆ E and 0 < ϕ < 1. It is noted that the main goal of sampling algorithms is to achieve a smaller subgraph similar to the original graph G. Sampling algorithms [8-10] play an important role in preprocessing, characterizing, studying, and analyzing the real networks. Network sampling can be applied for characterization of a small portion of a network subject to preserve main properties of the original network. For example, due to privacy restrictions of some OSN services, accessibility to the whole network information is not practical, therefore, the network is studied and characterized by estimated small samples from an important part of the original network. Taking into account large size, complexity, and dynamic properties of such real networks, most classic...

Balancing exploration and exploitation in memetic algorithms: A learning automata approach

“…Learning automata have a vast variety of applications in combinatorial optimization problems, computer networks, queuing theory, image processing, information retrieval, adaptive control, neural network engineering, cloud computing, social networks, and pattern recognition …”

“…In the latter case, the action probability vectors remain unchanged when the action taken is penalized by the environment. Learning automata have a vast variety of applications in combinatorial optimization problems, [25][26][27][28] computer networks, 25,[29][30][31][32][33][34] queuing theory, 35 image processing, 36 information retrieval, 37,38 adaptive control, 39 neural network engineering, 40,41 cloud computing, 42 social networks, 24,[43][44][45] and pattern recognition. 46…”

Channel estimation based on learning automata for OFDM systems

“…Thus, repeating of the process increases the possibility of selecting the ideal action. LA is used in several areas including wireless sensor networks, 24 online social networks, 25 resources allocation, 26 pattern classification, 27 signal processing, 28 and in some other troublesome areas of wireless communication such as those in the literatures. [29][30][31][32] Optimization problems become increasingly complex in the practical scenarios and real communication systems.…”

A novel organizing scheme of single topic user group based on trust chain model in social network

“…Therefore, user clustering cannot ignore the effectiveness of social relationships in information propagation. Many existing researches have already proposed that relationship among people exerts a key influence on information propagation . Also, users' interest, which leads people to cluster in user groups, influences the choice of each other by way of their trust relationships, indicating that relationship can be a critical factor in accurately organizing user groups in SNS.…”