ACO-based clustering for Ego Network analysis

摘要: Abstract The unstoppable growth of Social Networks (SNs), and the huge number connected users, have become these networks as one most popular successful domains for a large research areas. different possibilities, volume variety that SNs offer, has them an essential tool every-day working social relationships. One basic features any SN provides is to allow users group, organize classify their connections into groups, or “circles”. These circles can be defined using characteristics roommates, workmates, hobbies, professional skills, etc. problem finding taking account variety, dynamics important challenge wide Computer Science areas, Big Data, Data Mining Machine Learning among others. Problems related pre-processing, fusion knowledge discovering information from sources are still open question. This paper presents new Bio-inspired method, based on Ant Colony Optimization (ACO) algorithms, been designed find analyze circles. Given user in network, method able automatically determine compose his/her groups interest, so network will clustered components profiles dynamics. algorithm applied Ego where node centering (called “Ego”) represents being studied. In this work two ACO differ source used perform community tasks, designed. first uses extracted topology whereas second profile provided by users. proposed algorithms detect three Networks: Facebook, Twitter Google+. Finally, several databases previous SNs, experimental evaluation our methods carried out show how currently working.