A <i>k</i>-Hop Collaborate Game Model: Extended to Community Budgets and Adaptive Nonsubmodularity

Revenue maximization (RM) is one of the most important problems on online social networks (OSNs), which attempts to find a small subset users in OSNs that makes expected revenue maximized. It has been researched intensively before. However, exsiting literatures were based non-adaptive seeding strategy and simple information diffusion model, such as IC/LT-model. considered single influenced user measurement unit quantify revenue. Until Collaborate Game model appeared, it activity basic object compute An initiated by can only influence those whose distance are within k-hop from initiator. Based that, we adopt adaptive seed formulate Maximization under Size Budget (RMSB) problem. If taking into account product's promotion, extend RMSB Community (RMCB) problem, where be distributed over whole network. The objective function RMCB adatpive monotone not submodular, but some special cases, submodular. We study problem both speical submodular cases general non-submodular propose RMSBSolver RMCBSolver solve them with strong theoretical guarantees, respectively. Especially, give data-dependent approximation ratio for cases. Finally, evaluate our proposed algorithms conducting experiments real datasets, show effectiveness accuracy solutions.