Joint Colloquium Distinguished Lecture Series
Development as Design
Wednesday, October 20, 2010
My research is concerned with the design of appropriate and accessible information systems serving the needs of poor, indigenous, remote and otherwise marginalized communities in the developing and developed world. One major thread of this work is concerned with the design of accessible and trust-worthy applications for users with limited education, literacy skills and coming from diverse cultural and social backgrounds. I am also broadly interested in the impact that new kinds of data and communications tools can have for improving transparency and trust in governance, aid and philanthropy. My research group and I have recently developed "Avaaj Otalo", a phone-based voice message board allowing small farmers in India to ask and answer agricultural questions. Using any phone, farmers navigate a voice interface to record questions, obtain answers from experts, and to listen to and answer the questions of others. This system has been deployed in Gujarat, India for over eighteen months, consistently receiving hundreds of calls a week. We are currently planning a controlled, randomized evaluation of the impact of the system on agricultural productivity and income, in conjunction with planned implementations in several other states.
Another project, "Local Ground", is investigating the use of paper maps for documenting local geo-spatial knowledge. Previously untrained users annotate paper maps using cheap and accessible tools, such as colored markers and stamps, to provide contextualized knowledge about local events and spaces. This community-generated data is uploaded to our server, where the annotations are automatically extracted using a combination of simple computer vision and crowd-sourcing techniques. Overlaying this data on a Google Maps interface allows it to be compared to data obtained from other, more formal, sources.
Local Ground was recently used by low-income local teenagers from Richmond, California to participate in the planning of a public park in their community, culminated by presenting their ideas to the mayor's office. Drawing from these examples, I explore several themes in my work, including a) the design of cheap, "low-fidelity" interaction techniques allowing new populations to interact with and author content; b) the importance of "bottom-up" data for planning and evaluating development projects; and c) how "social data processing", interleaving automated and human-driven steps leveraging diverse incentive mechanisms and skills, can help bridge the potential gap between (a) and (b).
Tapan Parikh is a professor in the I-School, UC Berkeley. He received his M.S. and Ph.D. in Computer Science from the University of Washington, 1999 and 2007. His research interests include Human-Computer Interaction (HCI), Mobile Computing, Information and Communication Technologies for Development (ICTD), Information Systems for Microfinance, Smallholder Agriculture and Global Health.
|Return to EECS Joint Colloquium|