Semantic community detection using label propagation algorithm

Kianian, Sahar; Khayyambashi, Mohammad Reza; Movahhedinia, Naser

doi:10.1177/0165551515592599

Cited by 14 publications

(10 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wang and Fang used the SLTA to establish a user activity network through semantic data [12]. Kianian et al [13] proposed a semantic community discovery algorithm based on a label propagation algorithm. Based on the higher degree of intimacy between user nodes in social networks, the topic distribution is more similar [14,15].…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Tibetan Weibo User Group Division Based on Semantic Information in the Era of Big Data

Qiu

et al. 2018

Complexity

View full text Add to dashboard Cite

In the era of big data, group division in online social network analysis is a basic task. It can be divided into the group division based on static relationship and the group division based on dynamic relationship. Compared with the static group division, users express their semantic information in all kinds of social network behaviors, and they tend to interact with other users who have the same idea and attitude; this is how different groups are formed. In this paper, aimed at the issue that some Tibetan users use Chinese to publish microblogs on social platforms, a group division method based on semantic information of Tibetan users under the big data environment is proposed. When dividing a large number of Tibetan user groups in a social network, a large amount of semantic information of Tibetan users in the social network is first analyzed. Then, based on the semantic similarity between users, we aggregate the Tibetan users with high similarities into one group, thus achieving the final group division. The experimental results illustrate the effectiveness of the method of analyzing Tibetan user semantic information in the context of big data for group partitioning.

show abstract

Section: Related Workmentioning

confidence: 99%

“…(1) Enter the current network structure, take each node in the network as the initial group structure, and use (13) to calculate the similarity between each group structure.…”

Section: Group Division Algorithm For Tibetan Weibomentioning

confidence: 99%

Tibetan Weibo User Group Division Based on Semantic Information in the Era of Big Data

Qiu

et al. 2018

Complexity

View full text Add to dashboard Cite

show abstract

“…It iteratively updates seeds according to topological characteristics and propagates their membership degrees to non-seed nodes [14]. Kianian et al have analyzed semantic community detection using label propagation algorithm by focusing on user attributes [15]. Zhang et al have presented a memetic particle swarm optimization algorithm (MPSOA) to find the communities by hybridizing particle swarm optimization and tabu search for balancing the diversity and convergence [16].…”

Section: Related Workmentioning

confidence: 99%

Discovery of Multi-Objective Overlapping Communities within Social Networks using a Socially Inspired Metaheuristic Algorithm

Özbay¹,

Alataş²

2017

ijcna

View full text Add to dashboard Cite

-Frequently studied structural property of networks is community structure which is described as a group of users. User interactions inside the group are more than those outside the group. Communities in networks may be overlapped as users belong to multiple groups at once. This paper proposes a new socially inspired metaheuristic search and optimization algorithm, Parliamentary Optimization Algorithm (POA), to acquire promising solutions to overlapping community detection problems considering multiple objectives. The salient and unique feature of this work is that for the first time POA has been designed as a multi-objective search method for overlapping community detection. There is not any work about multi-objective overlapping community detection problem in the related literature. For this reason, simulation results of the proposed algorithm have not been compared with any results of works. The experimental studies on both artificial and real world social networks indicate that the POA ensures beneficial results for defining multi-objective overlapping community structure. A novel and interesting application area of POA has been introduced with this work. Parallel and distributed versions of social based POA with optimized parameters may also be efficiently designed and used for different social network problems.

show abstract

“…The mutation operator of real encoding is more complex than the binary encoding, and the common mutation operator design method includes two kinds of point type variation and uniform variation. Point mutation randomly choose a variant in a parent string, and then take a random number in [0,9] .Since the main purpose of the variation is to broaden your search solution to prevent falling into local optimal solution, so the real number coding variation in the integer part of multiple-choice position. 7 (2016) Image processing has three main stages: image fuzzification, modification of membership values, and, if necessary.…”

Section: Genetic Operators Designmentioning

confidence: 99%

“…The presence of these algorithms cause considerable disadvantage that edge positioning accuracy and scale edge detection problem i s a contradiction [8]; another traditional edge detection algorithm is also necessary to set a threshold value to distinguish noise point and edge point [9],but how to customize setting the optimal threshold adaptation is difficulty; Moreover, in different applications, people want to achieve different objectives, and therefore different definitions of edge and object edges. It is difficult to find a universal edge detection algorithm.…”

Section: Introductionmentioning

confidence: 99%