We are pleased to bring you this special issue in the ACM Transactions on Internet Technology (TOIT) titled Advances in Social Computing. Social computing is an emphasis area for TOIT and the current special issue follows an earlier one on the topic [Chopra et al. 2014]. Social computing is a growing interdisciplinary area of research. As editors, we cast our net wide, calling for contributions on themes such as models (e.g., interactions, organizations, and societies), methods (e.g., data mining, natural language, and social informatics), value (e.g., collective intelligence, trust, and privacy), and technology (e.g., cloud computing, the IoT, and wearables). The call attracted 35 fairly diverse submissions, each of which underwent rigorous peer review. We gratefully acknowledge the contribution of our many reviewers in providing timely and informative reviews. We list below the nine articles that made the cut and are included in this special issue. In Detecting Influencers in Multiple Online Genres, Rosenthal and McKeown present results on identifying influencers in online conversations by doing language analysis. Their methodology is informed by work in social science and based on identifying language patterns such as claims, agreements, argumentation, and so on. The authors report an extensive analysis of datasets from Twitter, LiveJournal, Create Debate, Wikipedia, and Political Forum demonstrating the wide applicability of their method. In Anonymous or Not? Understanding the Factors Affecting Personal Mobile Data Disclosure, Perentis et al. investigate the factors that influence a user into revealing personal information in mobile applications. Their methodology is based on conducting a participant study. Based on the study, they claim to be able to identify the key factors that influence users’ privacy behaviors. They also discuss the implications of their study for the design of more transparent Internet services. In Eliciting Structured Knowledge from Situated Crowd Markets, Goncalves et al. investigate the challenge of producing highly structured user responses, that is, knowledge, for arbitrary user questions. Their methodology is based on crowdsourcing, and it consists of not only getting answers from the crowd but also the criteria for evaluating the answers and a ranking for them. An interesting aspect of their approach is that they use situated (local) crowdsourcing, which they claim produces better information for context-dependent queries. In Formation and Reciprocation of Dyadic Trust, Roy et al. report a detailed empirical study of interpersonal trust in a multi-relational online social network. The
Unfortunately, ACM prohibits us from displaying non-influential references for this paper.
To see the full reference list, please visit http://dl.acm.org/citation.cfm?id=3080258.