Check out some of my ongoing research projects!

model-to-toca research agenda for human centered data science

I am deeply interested in methodological development at the intersection of HCI and data science. More broadly, I am very curious how people actually do or think or perceive practices of data analysis.

Currently, students are working on two different themes:

– what are the unanticipated outcomes of applying popular machine learning algorithms as black boxed systems? This is in response to certain recent developments and failures.

– survey research is one of the most popular approaches in doing hci research. In response to recent work on algorithmic gold standards, we ask, what is the general quality of survey research in hci and how can it be made better?

Representative Papers: GROUP’16,  JASIST’17, CSCW’17

keep-calm-and-quit-facebook-27understanding the spectrum of social media non use

I collaborate with Dr. Eric Baumer (assistant professor, Lehigh) and Prof. Geri Gay (professor, Cornell) to understand the different modalities and dynamics of social media non use.

Common understanding of social media non use is binary – either you use or you don’t.

However, in reality, this is more complicated. For instance, on Facebook, you can deactivate, delete, consider deactivating, consider deleting, experience non-volitionality of use, use, then not use, use, not use and then return to using and so on. Currently, students are building models to predict this spectrum of non-use and circumstances that lead to such behavior.

How do we examine these phenomena temporally?

Representative Papers: SM+S’15, CHB’16, CSCW’17

Bat-cave-check-insmodeling the dynamics of location disclosure in social networks

I am curious about the different norms and perceptions people have about sharing location in social networks.

Specifically, I am interested in privacy, surveillance, impression management and non-use in these settings.

Strategies that people employ to manage these concerns are also very exciting. For instance, obfuscation, deception and non use are some approaches that user commonly undertake in social networks.

Most recently, students have been working on understanding the dynamics of such aspects – for instance, we ask, is deception contagious?

Representative Papers: MobileHCI’13 , CSCW’15, Ubicomp’15

bd-sim-regdeveloping non-WEIRD models of privacy and security

I collaborate with Ishtiaque Ahmed (assistant professor, Toronto) and a wide group of other folks on broader issues encompassing privacy, security, access and surveillance in non-western, culturally situated and developing contexts.

For instance, recently, the Bangladesh government has instituted a mandatory biometric registration for anyone owning a mobile phone. This policy is a direct outcome of WEIRD governmental assumptions and policies yet is applied to a context which is very different. How do we think about their outcomes on computing and society?

Representative Papers: ICTD’16, CHI’17, ITID’17