Using Machine Learning to Understand the National Security Agency’s Data Surveillance Trends

• Author / Creator

  Damasah, Elliot Akwanfo

• For over six decades, the public has remained ignorant about the National Security Agency (NSA) and its activities and has been shielded from the agency’s invasive and unlawful projects. While the NSA’s activities have proven valuable to the United States and its allies, it sometimes undertakes projects that contravene human rights. Despite the power it possesses, knowledge about the NSA has remained scarcely accessible to the public. But the dynamics changed after Edward Snowden divulged and distributed NSA data to some journalists, giving us opportunities to explore the agency to increase our understanding of their work. As such, this study aims to add to the literature about the NSA by ascertaining the NSA’s discussions in SIDtoday, an online-based internal newsletter. It also aims to determine the trends of the various discussions within the agency and corroborate the agency’s discussions of data surveillance (an intentional observation of people as a means for collecting data).
  To test the hypothesis that the NSA discussed data surveillance issues, I collected text from the NSA’s SIDtoday newsletters and analyzed them with a machine learning algorithm called topic modeling. Topic modeling is a text analysis method capable of revealing themes within a large collection of text. The purpose was to identify discussions and trends in the collection of text. The results revealed that the NSA discussed eighteen (18) main topic areas, and included in these topics were Data Gathering and Target Analysis. Further analysis of these two topics indicated that the NSA engaged in data surveillance. Additionally, the trend analysis revealed the agency’s sustained and gradual increase in data surveillance from 2003 to 2006.
  From the results, it became evident that the NSA engaged in data surveillance. The other topics broached by the algorithm revealed the banal activities of the agency and its labor force. The trend analysis results affirm the agency’s inclination towards increased surveillance. Based
  ii
  on these findings, the public should be aware that the NSA is relentless in its efforts to undertake its surveillance responsibilities. That said, some people, especially activists and investigative journalists, ought to be vigilant in matters relating to their privacy.

• Subjects / Keywords

  • topic modeling
  • machine learning
  • NSA
  • national security agency
  • latent dirichlet allocation
  • trend analysis
  • data surveillance

• Graduation date

  Fall 2021

• Type of Item

  Thesis

• Degree

  Master of Arts

• DOI

  https://doi.org/10.7939/r3-w5fj-gt45

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.