Tracking Covert Groups in Twitter
Abstract
Tracking Covert Groups in Twitter Many covert and terror groups have a strong presence in social media. ISIS, in particular, is well known for its use of social media. Twitter is one of the dominant media used. Members of ISIS, ISIS sympathizers, and those watching ISIS, both women and men, send tweets about concerns, activities, and attitudes. Can such information, can such networks of people and topics, be used to identify the members of the group and potential converts? The ability to use social media and open source public information to identify covert networks and potential recruits could provide a critical new capability to the DoD. Having such information is critical from several perspectives. Such capabilities will increases situational awareness and provide a quantitative datasource for the development of metrics or measures of effectiveness. For example, change in the size of the observed potential converts could indicate the effectiveness of various campaigns aimed at “breaking” the terrornetwork. As another example, tracking both the network of actors and ideas, enables measuring changes in activity level and changes in hostility, commitment, and intent. This proposal is to develop the algorithms, techniques and procedures for identifying the members of a covert group of interest, the possible converts, attitudes, issues of interest, and any mentioned activities using publicly available Twitter data. This system will be web?based, semi?automated, and employ semisupervised machine learning algorithms. The core of our approach involves snowball sampling through the twitter API, high?dimensional network analytics for data reduction, and machine learning techniques across a high dimensional data array consisting of semantic and social network data. The proposed system will be tested using current and historic tweets. The covert group that this will be tested on is ISIS. Key challenges to be addressed include big data network metrics that go beyond volumetrics, multi?lingual data, biases in twitter usage, technology and users that impact the validity of various statistical techniques, and membership identification. Military Relevance: Social media informatics is key to open source exploitation. The proposed work will lead to new technology, procedures, and application scenarios for using social media to improve situational awareness, and gather critical intelligence. The search and identify technology will be useful in supporting counter?terrorism activities, and intelligence operations. Scientific Impact: This work creates a new dynamic network methodology that will enable new research to be conducted group participation, recruitment, and propaganda. This work also supports the development of “big data” metrics for social network research that go beyond the current volumetrics approach to indepth understanding of the underlying structure used by groups to generate cohesiveness and recruit new members. PI: Kathleen M. Carley, Carnegie Mellon University Requested Total Funds: $60,000. 1 year effort
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 08, 2016
- Source ID
- N000141512564
Entities
People
- Kathleen Carley
Organizations
- Massachusetts Institute of Technology
- Office of Naval Research
- United States Navy