Social media and networking platforms have billions of active users and leverage significant impacts on society. New types of social media and networking platforms or new features of existing platforms continue to be developed to meet users’ demands. With an increasingly large amount of unstructured social data on these platforms, social media and networking analytics research has the following scientific challenges:

  1. Difficult to analytically assess collective impact of social media and networking on societal polarization and other social phenomenon due to fragmentation of debates and discussions in the existing social media and networking platforms
  2. Lack of a central social media and networking platform for debate and discussions on important issues at national and international levels
  3. Hard to detect mis/dis-information, how it disseminates, and assess its impact
  4. Mining and using social media/networking data for logistics planning for disaster response
  5. Content-based indexing of unstructured and multimedia data on social media platforms and towards an integration with decision-making systems through deep learning methods
  6. Arduous to visualize large social network data


  1. Mining cyber argumentation data for collective opinions and their evolution
    • Faculty Lead: Susan Gauch
    • Objectives
      • Develop a cyber discourse social network platform
      • Collect data using the developed cyber discourse social network platform
      • Develop natural language processing algorithms to analyze discourse data collected by the platform and existing data
  2. Socio-computational models for safer social media
    • Faculty Lead: Nitin Agarwal
    • Objectives
      • Characterize online information environment (OIE)
      • Develop socio-computational models to identify key actors and key groups of actors
      • Study tactics, techniques, and procedures (TTPs) of deviant cyber campaigns
      • Develop socio-computational models to measure power of a cyber campaign
  3. Auto-annotation of multimedia data
    • Faculty Lead: Ashlea Bennett Milburn
    • Objectives
      • Develop multimedia indexing methods for social media data
      • Design and implement deep learning methods for multimedia data
      • Build Integrated smart applications based on unstructured multimedia data
  4. Informing disaster response with social media
    • Faculty Lead: Ashlea Bennett Milburn
    • Objectives
      • Extract and index content describing transportation infrastructure status from social platforms
      • Fuse data from social platforms describing transportation infrastructure status with other data sources
      • Assess credibility of data inputs from the objectives within this goal
      • Develop routing algorithms that use inputs from this goal’s objectives to support routing for disaster response

Advancing the State of Knowledge

Envisioned advancements to the state of the knowledge by each goal are as follows:

  • Goal 1 will advance the state of the knowledge in argumentation polarization modeling by developing innovative quantitative opinion polarization techniques to model the formation and evolution of opinion polarization in large-scale cyber argumentation and deliberation with social networks.
  • Goal 1 will elevate methods and techniques to predict individual or collective opinions on single or multiple solutions of issues using collaborative filtering and machine learning-based techniques.
  • Goal 1 will improve social network methodology and social network sampling and data generation.
  • The proposed research in Goal 2 advances our understanding of the role of Information and Communication Technology-mediated communications in the formation of emergent organizations with implications to business, marketing (explaining viral behaviors), and many other settings.
  • Goal 2 is of particular interest to information scientists exploring the influence of social systems on user behaviors; studying ties between people, technology, and institutions; examining organizational structures, roles, and crowd processes; investigating the notions of individual and collective identities in a variety of information systems supporting crowdsourcing, citizen participation, eGovernance, crisis and disaster management, and several other manifestations of emergent organizations.
  • Goal 2 would contribute to the theory of collective action to model the dynamics of deviant cyber behaviors, borrow from the literature on collective identity formation to explain the motivation needed to sustain such coordinated acts, assimilate factors pertaining to collective failures/success, and leverage notions of hypergraph to model complex (multidimensional and supra-dyadic) relations commonplace among members of deviant groups.
  • The proposed research in Goal 2 advances the literature on cyber-collective actions and study the role of social media in organization and coordination of cyber social movements from individual, community, inter-organizational, and transnational perspectives.
  • The proposed research in Goal 2 develops socio-computational predictive models that are efficient, reliable, scalable, explainable, reproducible, and theoretically grounded to help understand behaviors from social media platforms.
  • Goal 3 will allow decision making processes to utilize multi-source, heterogenous and multimodal data towards better performance, as well as expand the scope of learning and artificial intelligence techniques to multimedia data.
  • Goal 3 will afford applications such as disaster recovery to take advantage of vast, increasingly popular multimedia data that is often unstructured and insufficiently indexed.
  • Goal 4 provides methods for indexing and fusing transportation infrastructure status data from a variety of sources (e.g., social media, satellite imagery, traffic camera videos).
  • Goal 4 applies credibility detection methods for transportation infrastructure status data on social media.
  • Goal 4 enables the use of near-real-time transportation infrastructure status data in logistics planning methods to support disaster response.
  • Goal 4 also creates new vehicle routing models and solution approaches for complex real-time logistics planning problems in disaster response.