On Trust, Privacy, and Misinformation in Health Social Media

  • Author / Creator
    Samuel, Hamman W
  • It has gotten increasingly harder for laypersons to determine the veracity of online health information. This is because of the explosion of content in health social media, allowing anyone with an Internet connection to create and propagate health-related content. This includes both innocuous and malignant pseudo-medical advice. On the other hand, medical professionals are able to discern medical facts from fiction using systematic methodologies. This thesis develops and evaluates pragmatic computational models for evaluating the veracity of health-related online content. Firstly, medical knowledge and evidence-based practices are incorporated into health social media through the MedFact algorithm and Veracity Score. Secondly, privacy and anonymity requirements of social media are taken into account using the Iron Mask algorithm and Trust-Preserving Pseudonyms. Thirdly, these solutions are incorporated into a health portal for patients and medics, code-named Cardea, that models various types of interactions occurring on health social media, including real-time chat rooms, blogs, question-answering, and support groups. Cardea allows users to share experiences, ask questions, and get answers in three streamlined environments: Patient to Patient, Patient to Medic, and Medic to Medic. Patients are able to chat with other patients, create support communities, and also ask questions specifically to medical experts. Medics can respond to patient questions and also have private and secure discussions with other medics. The road map for this manuscript is organized by chapters into three predominant groups. Chapters 1 and 2 provide the background to the thesis. In Chapter 1, the motivation, background, and thesis statement are provided. In Chapter 2, a survey of literature related to trust, privacy, and evidence-based medicine is covered. Chapters 3, 4, and 5 expand the key concepts of the hypotheses on trust, privacy, and health social media. In Chapter 3, the MedFact algorithm is explicitly defined as an objective metric for computational estimation of trust. In Chapter 4, the Iron Mask algorithm is explained as a mechanism towards preventing social stigma while preserving reputation. In Chapter 5, Cardea is described in detail, including its embedded frameworks and components of trust, privacy, and security. Chapters 6, 7, and 8 cover additional artifacts developed as part of this research for use within Cardea for data collection, content recommendation, and duplicate content detection. Chapter 9 concludes this manuscript with an outlook on future research potential. Takeaway boxes are used throughout the manuscript to highlight and summarize key concepts, results, and contributions. Ultimately, this thesis gives new perspectives on a computational definition of trust with an awareness of privacy in health social media. The proposed methods have the potential for assisting users to sift through large volumes of online information and make informed decisions about their health using trustworthy information sources without compromising privacy.

  • Subjects / Keywords
  • Graduation date
    Spring 2021
  • Type of Item
  • Degree
    Doctor of Philosophy
  • DOI
  • 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.