Neil Patel
Department of Computer Science, Stanford University
Sharing Information in Rural Communities Through Voice Interaction

You need Flash player 8+ and JavaScript enabled to view this video.

Online communities enable people to access and share information, but they are out of reach for poor and isolated communities around the world. Mobile phones have the potential to overcome the PC's accessibility, affordability, and familiarity barriers. However, most mobile information services limit rural populations to being passive knowledge consumers, not active producers. My dissertation explores the design and usage of voice-based social media for rural communities. My collaborators and I designed Avaaj Otalo ("voice stoop") a voice message board application that allows small-scale farmers in India to share agricultural advice by posting, listening to, and replying to others' messages using any phone. AO has been live and accessed by thousands since 2009. We have deployed voice message boards with five other partners across India working in agriculture, labor rights, women's empowerment, and education, logging over 30,000 calls.

This talk presents research guided by three essential questions for designing systems in this context: how to design effective UIs for navigating audio content; what community dynamics emerge; and how to support sustained user engagement. Prior work has assumed that spoken input is most effective for technology novices with limited literacy. We tested this hypothesis in controlled and natural settings, instead finding that touchtone input was more effective and preferable. Next, we tested whether information on AO was more influential when it came from nominal authorities compared to peers. Contrary to stated preference, participants acted more upon the same information when it came from a peer. Finally, we tested various financial incentives and motivational messaging to elicit usage. Free calls increased usage and posting, but so did free access to just selected content and message recording. Phone broadcasts with twenty-second motivational messages increased both calling and posting.

Neil Patel is a PhD student in Computer Science at Stanford University, where he works in the HCI group. His research explores the design and usage of ICTs for underserved populations; specifically, he works on voice-based social media for rural communities in India. Prior to graduate school, Neil had some of the best times of his life studying Computer Science and Business Administration at UC Berkeley. He currently commutes between his homes in California and Ahmedabad, Gujarat.