I'm a 3rd year Ph.D. student at Northwestern University in the Technology and Social Behavior (TSB) program advised by Professor Haoqi Zhang and Professor Darren Gergle. My research interests broadly falls at the intersection of human-computer interaction (HCI), social and crowd computing, and artificial intelligence.
My current work studies the concept of networked orchestration--how work organizations and learning communities disperse the control of orchestrating learning and support across a network of resources--and the design of networked orchestration technologies that can support communities in more effectively conducting networked orchestration. I hope that these technologies will enhance the ways we collaborate with each other in our organizations and communities.
From our prior study of Networked Orchestration in the Design, Technology, and Research (DTR) program, we found that students struggled to identify their planning and help-seeking needs, requiring mentors to use their limited time to conduct time-consuming diagnosis to identify those needs. However, that also restricts which student needs they can address, resulting in some skills that are essential for effectively working in networked orchestration being under-practiced. Furthermore, the existing scale already makes it difficult for mentors to quickly recall previously discussed issues and use that information to facilitate further diagnosis with students.
To address these issues, we are studying the design of orchestration scripts: a platform for encoding mentoring strategies for identifying and addressing student needs into a machine-executable form that can be continually enacted in the community without additional burden on the mentor. Orchestration scripts enable the direct surfacing of students' needs monitorable conditions, rather than just providing high-level awareness from which needs must still be determined. Further, orchestration scripts allow for the encoding and enacting of complex strategies through actionable feedback that leverage multiple resources available in the DTR community.
Orchestrating learning and support for working on complex, ill-structured problems is often hindered in its efficacy and scale by practical orchestration challenges, including limited expert resources for supporting many novices simultaneously. Here, networked orchestration–the dispersion of orchestration control across a community–is necessary for successfully responding to novices’ many needs.
To understand the practices and challenges of networked orchestration, we conduct a qualitative study of the Design, Technology, and Research (DTR) program for undergraduate research training. Through in-depth interviews and participant observations of DTR, we develop an emergent framework for how students and mentors address students’ needs using community resources, and how mentors coach students on improving their work practices. We also identify students’ struggles in seeking help from unfamiliar resources, and mentors’ struggles in diagnosing and tracking the many needs of their students, and use these to motivate the design of future networked orchestration technologies.
Participatory sensing systems must account for both the needs of data contributors and the data collection goals. Existing approaches tend to emphasize one or the other, with opportunistic and directed approaches making opposing trade-offs between providing convenient opportunities for contributors and collecting high-fidelity data. Instead, we study a new approach called 4X that leverages different data collection mechanisms (i.e., opportunistic or directed collection) depending on where people are, what data is available, and what people care about by using previously collected data to draw people to places they may care about.
To study the efficacy of 4X, we implemented 4X into LES, an application for collecting information about campus locations and events. Results from two field deployments (N = 95, N = 18) show that the 4X framework created 34% more opportunities for contributing data without increasing disruption, and yielded 49% more data by directing users to locations of interest.
Jupyter Notebooks, Pandas, Numpy, Scikit Learn, R