Physical events happening in the real world usually trigger reactions and discussions in the digital world; a world most often represented by Online Social Media such as Twitter or Facebook. Mining these reactions through social sensors offers a fast and low cost way to explain what is happening in the physical world. A thorough understanding of these discussions and the context behind them has become critical for many applications like business or political analysis. This context includes the characteristics of the population participating in a discussion, or when it is being discussed, or why. As an example, we demonstrate how the time of the day affects the prediction of traffic on highways through the analysis of social media content. Obtaining an understanding of what is happening online and the ramifications on the real world can be enabled through the automatic summarization of Social Media. Trending topics are offered as a high level content recommendation system where users are suggested to view related content if they deem the displayed topics interesting. However, identifying the characteristics of the users focused on each topic can boost the importance even for topics that might not be popular or bursty. We define a way to characterize groups of users that are focused in such topics and propose an efficient and accurate algorithm to extract such communities. Through qualitative and quantitative experimentation we observe that topics with a strong community focus are interesting and more likely to catch the attention of users.
Consequently, as trending topic extraction algorithms become more sophisticated and report additional information like the characteristics of the users that participate in a trend, significant and novel privacy issues arise. We introduce a statistical attack to infer sensitive attribute values of Online Social Networks users that utilizes such reported community-aware trending topics. Additionally, we provide an algorithmic methodology that alters an existing community-aware trending topic algorithm so that it can preserve the privacy of the involved users while still reporting trending topics with a satisfactory level of utility. From the user’s perspective, we explore the idea of a cyborg that can constantly monitor its owner’s privacy and alert them when necessary. However, apart from individuals, the notion of privacy can also extend to a group of people (or community). We study how non-private behavior of individuals can lead to exposure of the identity of a larger group. This exposure poses certain dangers, like online harassment targeted to the members of a group, potential physical attacks, group identity shift, etc. We discuss how this new privacy notion can be modeled and identify a set of core challenges and potential solutions.
