Mohammed N. Alandoli, M. Shehab, M. Al-Ayyoub, Y. Jararweh, Mohammad Al-Smadi
{"title":"利用gpu加速社交网络中基于fcm的社区检测","authors":"Mohammed N. Alandoli, M. Shehab, M. Al-Ayyoub, Y. Jararweh, Mohammad Al-Smadi","doi":"10.1109/CSIT.2016.7549467","DOIUrl":null,"url":null,"abstract":"One of the important features of Social Networks (SNs) is community structure detection. Several methods have been proposed to address this problem. One of the interesting methods is based on the famous Fuzzy C-Means (FCM) clustering algorithm. This method consists of three phases: spectral mapping, FCM clustering and modularity computation. Despite being very effective, this method is actually inefficient to deal with large-scale networks. A parallel implementation using GPUs is one of the feasible solutions to address this problem. Hence, this research presents a parallel implementation of FCM and modularity components of the algorithms. The implementation follows the hybrid CPU-GPU approach. We study the many factors affecting the performance speedups, such as the number of dimensions/features and the network size.","PeriodicalId":210905,"journal":{"name":"2016 7th International Conference on Computer Science and Information Technology (CSIT)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2016-07-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"16","resultStr":"{\"title\":\"Using GPUs to speed-up FCM-based community detection in Social Networks\",\"authors\":\"Mohammed N. Alandoli, M. Shehab, M. Al-Ayyoub, Y. Jararweh, Mohammad Al-Smadi\",\"doi\":\"10.1109/CSIT.2016.7549467\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"One of the important features of Social Networks (SNs) is community structure detection. Several methods have been proposed to address this problem. One of the interesting methods is based on the famous Fuzzy C-Means (FCM) clustering algorithm. This method consists of three phases: spectral mapping, FCM clustering and modularity computation. Despite being very effective, this method is actually inefficient to deal with large-scale networks. A parallel implementation using GPUs is one of the feasible solutions to address this problem. Hence, this research presents a parallel implementation of FCM and modularity components of the algorithms. The implementation follows the hybrid CPU-GPU approach. We study the many factors affecting the performance speedups, such as the number of dimensions/features and the network size.\",\"PeriodicalId\":210905,\"journal\":{\"name\":\"2016 7th International Conference on Computer Science and Information Technology (CSIT)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2016-07-13\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"16\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2016 7th International Conference on Computer Science and Information Technology (CSIT)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/CSIT.2016.7549467\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2016 7th International Conference on Computer Science and Information Technology (CSIT)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CSIT.2016.7549467","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Using GPUs to speed-up FCM-based community detection in Social Networks
One of the important features of Social Networks (SNs) is community structure detection. Several methods have been proposed to address this problem. One of the interesting methods is based on the famous Fuzzy C-Means (FCM) clustering algorithm. This method consists of three phases: spectral mapping, FCM clustering and modularity computation. Despite being very effective, this method is actually inefficient to deal with large-scale networks. A parallel implementation using GPUs is one of the feasible solutions to address this problem. Hence, this research presents a parallel implementation of FCM and modularity components of the algorithms. The implementation follows the hybrid CPU-GPU approach. We study the many factors affecting the performance speedups, such as the number of dimensions/features and the network size.