Social media is changing the way we access information, share experiences, and in general, interact with others. Recent statistics show that approximately 20 percent of time spent on the Web involves social media. While some of this time is well spent (enjoying all its benefits), some of the time is spent trying to overcome the challenges that social media has introduced, such as identifying credible information, searching for relevant content, ensuring that privacy constraints are met, and overcoming security breaches.
A currently popular form of social media, microblogging, enables individuals to share small pieces of information with many others whom they might barely know. These individuals often have minimal credentials — sometimes not even a full name. It isn’t clear whether they created their own credentials and, if not, what the origin of those credentials might be. Without knowing this, it’s difficult to judge the credibility of information that microbloggers provide. Currently, users deal with this by searching through other media to check whether the content is actually true or try to identify ways to cross-check the information with other users. It would be tremendously important if agents could help identify trustworthy users in the system, as well as provide ways to reason about the information’s provenance. Even when the information’s credibility is guaranteed, there’s too much information to be processed manually. Ideally, it would be helpful to have agents that can mine social media, differentiate between important and unimportant information, and communicate their results clearly.
A side effect of sharing information over social media is that users’ privacy is violated easily, either because shared content reaches an unintended audience or other users share content about individuals without the individual’s consent. Managing an individual’s privacy in social media would be eased if personal agents could help maintain requirements, verify whether the system abides with the requirements, and cooperate with other agents as needed to ensure that the requirements are met.
While most current social media is intended for sharing content, future social media applications could offer models for other forms of interactions, including business and government. Such models could make use of agents that form teams, partnerships, and communities, foster communications, and collaborate to formulate policies and reach decisions. To realize these new forms, underlying computational challenges would need to be addressed.
This special issue will address the questions, challenges, and opportunities that arise at the intersection of agents and social media. These contributions can include theoretical and applied research related to the modeling, design, and development of agents and multiagent systems for social media. Relevant topics include (but aren’t limited to) the following:
decision making in social media (application of agreement technologies, negotiation, and argumentation);
mining social media (data analytics and visualization);
searching (information retrieval, information fusion, and crowdsourcing);
personalizing social media (including the content and view);
ensuring cybersecurity for social media (surveillance, intrusion detection, vulnerability analysis, and information forensics);
respecting privacy in social media (learning users’ privacy constraints, detecting privacy violations, and negotiating privacy constraints on mutual posts);
engendering trust based on provenance for social media;
conducting business over social media (models, teamwork, gamification, and agent-human collectives);
playing games over social media (interaction design, social games, and augmented reality);
forming communities over social media (building, detecting, and maintaining); and
engaging the elderly.