Banner Image


Always curious, we find research work to be super exciting and like to support it as much as we can. Our research with NGOs like CGNetSwara and private institutions like Voicedeck Pvt Ltd has been published in renown journals across the world.

Featured Publications:

A Hybrid Multi-Modal System for Conducting Virtual Workshops Using Interactive Voice Response and the WhatsApp Business API

CHI EA 2021: Extended Abstracts of the 2021 CHI Conference on Human Factors in Computing Systems

Interactive voice response (IVR) forums such as CGNet Swara and Avaaj Otalo have played a pivotal role in empowering marginalised communities by providing an avenue to make their voice heard through simple phone calls. At the same time, growing internet penetration and affordable data plans are altering the ways in which rural Indian communities access and consume information. Within the context of a shift from voice to richer content environments, we present the design of a multi-modal awareness generation and data collection platform built around IVR and the WhatsApp Business API. This model was deployed for delivering virtual training modules to cotton farmers in rural Maharashtra. During the 27 day deployment, 176 people participated in the intervention, out of which 122 and 54 attempted the modules on IVR and WhatsApp, respectively. In this paper, we highlight some of the interesting findings and lessons learnt during the intervention.

Read Full PaperDownload PDFMay 08, 2021

Demo: A WhatsApp Bot for Citizen Journalism in Rural India

COMPASS 2021: ACM SIGCAS Conference on Computing and Sustainable Societies

Increasing penetration of Internet-enabled smartphones in low-resource areas makes them an attractive platform for engaging emerging users. In this paper, we demonstrate how a voice forum for citizen journalism in rural India– previously accessible via an Interactive Voice Response (IVR) system– can be naturally supported and enriched using a chatbot. Implemented using the WhatsApp Business API, the bot enables submission of both audio (with or without image) and video stories. Following review by moderators, stories are published on a website and social media sites, and can also be browsed interactively using the WhatsApp bot. This multi-way, intermediated model of communication expands the scope and functionality of typical WhatsApp groups while offering significant cost savings relative to IVR systems. In the first 9 weeks of a long-term deployment, the bot demonstrated high usability and acceptance and resulted in 218 published stories from 27 users.

Read Full PaperDownload PDFMay 23, 2021

Learnings from Technological Interventions in a Low Resource Language: A Case-Study on Gondi

LREC (The International Conference on Language Resources and Evaluation) 2020

The primary obstacle to developing technologies for low-resource languages is the lack of usable data. In this paper, we report the adoption and deployment of 4 technology-driven methods of data collection for Gondi, a low-resource vulnerable language spoken by around 2.3 million tribal people in south and central India. In the process of data collection, we also help in its revival by expanding access to information in Gondi through the creation of linguistic resources that can be used by the community, such as a dictionary, 'children's stories, an app with Gondi content from multiple sources and an Interactive Voice Response (IVR) based mass awareness platform. At the end of these interventions, we collected a little less than 12,000 translated words and/or sentences and identified more than 650 community members whose help can be solicited for future translation efforts. The larger goal of the project is collecting enough data in Gondi to build and deploy viable language technologies like machine translation and speech to text systems that can help take the language onto the internet.

Read Full PaperDownload PDFMay 11, 2020

Facilitating Media Distribution with Monetary Incentives

CHI (Conference on Human Factors in Computing Systems) 2020

Community media forums such as CGNet Swara and Mobile Vaani leverage interactive voice response (IVR) technology to enable feature phone owners in underprivileged regions of the world to call a toll-free phone number and listen to and report local stories. This excludes potential users living in areas without any mobile network. In this paper, we describe the deployment of an app that both facilitates a new means of media dissemination using a novel incentive scheme, and enables content collection, from no-network areas. In partnership with CGNet Swara, we deployed our app in the rural Indian states of Chhattisgarh and Telangana. In one month, 20,955 stories were transferred through Bluetooth to 2,443 unique phones, for which a total of $930 in mobile airtime credits was disbursed to 307 of the 680 users that installed the app. In an 81-day period, 537 local stories were reported using the app from 117 unique users. We present a quantitative and qualitative analysis of user behavior.

Read Full PaperDownload PDFApril 29, 2020

Learnings from an Ongoing Deployment of an IVR-based Platform for Voter Awareness

ACM CSCW (Conference on Computer-Supported Cooperative Work and Social Computing) 2019

Spreading awareness among voters is a critical first step towards ensuring democratic participation. We report results from an ongoing deployment in India where we use an Interactive Voice Response (IVR) system to raise awareness about voting. People can call in to the number and answer a few questions on voting. They receive a mobile airtime top-up if they answer correctly, and an explanation of the correct answer if they do not. The system also serves as a survey instrument and we use it to collect data on the percentage of people who have their voter ID cards that enable them to vote. Extending past work on IVR-based mass awareness, we employ a different way of presenting content---a quiz instead of a tutorial---and of incentivizing usage---no monetary incentive for referrals except for specified referrers. In 24 days of deployment, over 1900 people have called the system, out of which 1245 answered all questions correctly in their first attempt, and 234 answered correctly after learning from their initial incorrect answers. We present a quantitative and qualitative analysis of user behavior.

Read Full PaperDownload PDFNovember 9, 2019