پديد آورنده :
اسلامي، تابان
عنوان :
الگوريتمي جديد، مقياس پذير و مقاوم براي بيشينه سازي تاثير در شبكه هاي اجتماعي
مقطع تحصيلي :
كارشناسي ارشد
گرايش تحصيلي :
كامپيوتر - هوش مصنوعي
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر
صفحه شمار :
ده، 74 ص.: مصور، جدول، نمودار
استاد راهنما :
محمدحسين سرايي
توصيفگر ها :
مدل انتشار , بهينه سازي چندهدفه
تاريخ نمايه سازي :
1394/05/28
استاد داور :
عبدالرضا ميرزايي
تاريخ ورود اطلاعات :
1396/10/04
رشته تحصيلي :
برق و كامپيوتر
دانشكده :
مهندسي برق و كامپيوتر
چكيده فارسي :
به فارسي و انگليسي
چكيده انگليسي :
75 A New Scalable and Robust Algorithm for Influence Maximization in Social Networks Taban Eslami t eslami@ec iut ac ir Date of Submission Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language FarsiSupervisor MohammadHossein Saraee Saraee@cc iut ac irAbstractSocial networks are sets of individuals who communicate to each other based on a specificpurpose depending on the type of social network This communication may be friendship business or scientific cooperation or other kinds of relations Nowadays social networks arevery popular in the entire world and this popularity makes them suitable for large scaleadvertisements Relationships between different users in a social network increase theamount of influence on each user such that related users will have similar interests andactivities One of the important problems in the area of social network is influencemaximization problem i e identifying a set of key nodes in a social network that maximizesthe influence spread This problem has gained tremendous attention in recent years Oneapplication of such problem is time bounded influence maximization for viral marketing While a number of algorithms exists that give satisfactory performance for influencemaximization in large networks time bounded influence maximization still remains an openproblem In this thesis we introduce a new multi objective optimization based approach forinfluence maximization considering two objectives maximizing influence and minimizingdiffusion time We adapted NSGA II algorithm and in order to make the running time ofthe algorithm feasible for optimization over large networks we developed two heuristics forthe computation of influence spread and diffusion time of sets of users In our algorithm atthe end of the optimization phase the influence of resulting sets of nodes from NSGA IIalgorithm is evaluated at constrained time bounds and the final optimal set of nodes underimposed time bounds are given as output which gain higher influence compared to existingalgorithms at specific time bounds Keywords Influence maximization Diffusion model Multi objective optimization Socialnetworks
استاد راهنما :
محمدحسين سرايي
استاد داور :
عبدالرضا ميرزايي
