click here to download the base paper of social networks
click here to download the abstract project of social networks
ABSTRACT
The cascading of sensitive information such as private contents and rumors is a severe issue in online social networks. One approach for limiting the cascading of sensitive information is constraining the diffusion among social network users. However, the diffusion constraining measures limit the diffusion of non-sensitive information diffusion as well, resulting in the bad user experiences. To tackle this issue, in this paper, we study the problem of how to minimize the sensitive information diffusion while preserve the diffusion of non-sensitive information, and formulate it as a constrained minimization problem where we characterize the intention of preserving non-sensitive information diffusion as the constraint. We study the problem of interest over the fullyknown network with known diffusion abilities of all users and the semi- known network where diffusion abilities of partial users remain unknown in advance. By modeling the sensitive information diffusion size as the reward of a bandit, we utilize the bandit
framework to jointly design the solutions with polynomial complexity in the both scenarios. Moreover, the unknown diffusion abilities over the semi-known network induce it difficult to quantify the information diffusion size in algorithm design. For this issue, we propose to learn the unknown diffusion abilities from the diffusion process in real time and then adaptively conduct the diffusion constraining measures based on the learned diffusion abilities, relying on the bandit framework. Extensive experiments on real and synthetic datasets demonstrate that our solutions can effectively constrain the sensitive information diffusion, and enjoy a 40%less diffusion loss of non-sensitive information comparing with four baseline algorithms
INTRODUCTION
The prevalence of online social networks such as Facebook, Twitter and Wechat facilitates the information diffusion among users, and thus enables the efficient promotion of positive information’s, e.g., products, news, innovations . Although such efficient diffusion can easily lead to large- scale diffusion called information cascading, the unconstrained cascading behavior could meanwhile cause the sensitive information to be incautiously diffused over the network . Here the sensitive information refers to any kind of information that needs to be prohibited from cascading such as rumors, personal contents, and trade secrets. The cascading of such sensitive information may cause the risk of leaking users’ privacies or arising panics among publics . With this concern, several social network medias (e.g., Facebook, Twitter) have claimed authorities to block accounts of users and delete some posts or tweets when they violate relevant rules about privacies or securities . Thus network managers are able to take measures to prohibit the cascading of sensitive information. The existing attempts that share the closest correlation with prohibiting sensitive information diffusion belong to the rumor influence minimization, whose current strategies can mainly be classified into two aspects. The first is diffusing the truths over network to counteract rumors . However, diffusing truths is only suitable for constraining the rumors, while is not
suitable for constraining the diffusion of the other kinds of sensitive information’s, including personal information’s, trade secrets, and etc. The second is temporarily blocking a number of users with high diffusion abilities or blocking a number of social links among users in hope of minimizing the diffusion of a rumor. Although such strategy is effective for preventing rumors about some significant events like earthquakes, terrorist attacks and political elections, it is unrealistic for network managers to adopt this strategy on constraining the diffusion of sensitive information’s with various contents that widely exist in our daily lives. If network managers take such measure, it is required to block a much larger size of users or links. Then two critical problems arise. Firstly, blocking too many users or social links will degrade user experiences and may arouse complaints for the right violation. Secondly, blocking users or social links for
restraining rumors also brings the loss of the diffusion of positive information’s, say information loss, which is not beneficial to the viral marketers that utilize information cascading to promote products.
Introduction
EXISTING SYSTEM
The existing attempts that share the closest correlation with prohibiting sensitive information diffusion belong to the rumor influence minimization, whose current strategies can mainly be classified into two aspects.
- The first is diffusing the truths over network to counteract rumors. However, diffusing truths is only suitable for constraining the rumors, while is not suitable for constraining the diffusion of the other kinds of sensitive informations, including personal informations, trade secrets, and etc.
- The second is temporarily blocking a number of users with high diffusion abilities or blocking a number of social links among users in hope of minimizing the diffusion of a rumor.
PROPOSED SYSTEM - To tackle the above challenges, we utilize the constrained combinatorial multi-arm bandit framework to jointly design our solutions over the fully- known and semi-known networks, where we take the diffusion size of sensitive informations as the reward of a bandit and model the probability variations as the arms in bandit.
- With this mapping, we determine the probability variations through a constrained arms picking process with the aim of minimizing the obtained rewards.
- Through incorporating the constraint of diffusion probability variations into the construction of the arms of bandit, we relax the problem of interest into an unconstrained minimization problem when determining the diffusion probability variations based on the arms.
- This enables us to determine the probability variations via social 1.https://www.douban.com/ links with high efficiency.
- Furthermore, for coping with the unknown diffusion abilities over the semi known network, we propose to iteratively learn the unknown diffusion abilities through learning the reward distributions of the arms based on the rewards obtained from previously picked arms, and then determine the
diffusion probability variations based on the learned reward distributions of arms.
