Kai Wei



More Information


Doctoral student Kai Wei seeks to understand online social dynamics and approaches for characterizing the patterns. Wei’s current work focuses on characterizing the patterns of online prejudiced speech and its changes in the context of offline events. Her work in this area uses linguistic analysis, data mining, temporal and spatial analysis, and is being conducted in collaboration with researchers in the Pitt Computational Social Dynamics Lab (PICSO). Kai also has been actively involving in promoting interdisciplinary research between social work and information and computer science. In the past, she has co-chaired multiple sessions related to this effort including the opportunities and challenges of interdisciplinary research and the role of social media in prevention research.

Research Interests

  • Topic interests: Dynamics of prejudiced speech and related outcomes and interventions, social determinants of health
  • Methodological interests: Computational approaches for analyzing large-scale text data

Selected Publications

Booth, J., Wei, K., Little, A. (in press). Examining the Impact of Food Environment Changes on County-level Obesity Prevalence in the Appalachian Region.  Journal of Health Disparities Research.

Kyere, E., Joseph, A., & Wei, K. (in press) Alternative to Zero Tolerance Policies and Out-of-School Suspensions: A Multi-tiered Centered Perspective. Journal of Ethnic & Cultural Diversity in Social Work.

Booth, J., Chapman, D., Ohmer, M., Wei. K. (2018). Examining the Relationship between Level of Participation in Community Gardens and their Multiple Functions. The Journal of Community Practice, 26(1), 5-22. (Link)

Wei, K., & Booth, J. (2017). The Association between Neighborhood Factors and Mexican Americans’ Mental Health Outcomes: A Systematic Review. Journal of Sociology & Social Welfare, 44(2), 133-160. (Link)

Urbaeva, J., Booth, J., Wei, K. (2017). The Relationship between Cultural Identification, Family Socialization and Adolescent Alcohol Use among Native American Families. Journal of Child and Family Studies, 1-13. (Link)

Chung, W-T., Wei, K., Lin, Y.-R., Wen, X. (2016). The Dynamics of Group Risk Perception in the US After Paris Attacks. In Proc. of the 8th International Conference on Social Informatics. (Link) [Full paper; Acceptance rate: 28%; Best Paper Award]

Wei, K., & Lin, Y. R. (2016). The evolution of Latino threat narrative from 1997 to 2014. IConference 2016 Proceedings. (Link[Best Poster Paper Finalist]

Booth, J., & Wei, K. (2015). The Latino health paradox: Examining the Mexican American Experience. Contemporary issues for people of color: Surviving and thriving in the U.S. today, Volume 4: Health and wellness, Greenwood.

Current Projects

Using social media to evaluate grassroots campaigns in real-time. This study aims to provide an analytical pipeline for evaluating grassroots campaigns using social media data. Engaging in grassroots campaigns is one way for citizens to promote a more just and equitable society. Evaluation of grassroots campaigns in real-time can provide immediate feedback for activists and community organizers. I recently presented preliminary findings in a  symposium, where I described an analytical pipeline that uses a quasi-experimental study design and text mining techniques, and demonstrated its effectiveness by examining changes in Twitter users’ perception of fairness in immigration discussions following the #NoBanNoWall campaign.

Understanding the language of human bias in social media. This aims to quantify human bias using social media data. Human bias refers to people's one-sided point of view about a person or a group, which can be favor or disfavor one thing, person, or group compared with another. Given that language and cognition are closely related, we argue that people’s language use can be used as an indicator for measuring people’s cognitive bias. This work is being conducted in collaboration with Mengdi Wang in Intelligent Systems Program, and Dandi Zhao and Chaitanya Parsana in School of information and Computing.