ICTD · 2020
DIA: A Human AI Hybrid Conversational Assistant for Developing Contexts
Proceedings of the 2020 International Conference on Information and Communication Technologies and Development (ICTD), Article 24, 1-5.
About the research
Abstract
Social media messaging applications such as WhatsApp and Facebook Messenger reached 2.3 billion users in 2019, with the majority of users emerging from developing countries. This high usage opens possibilities for designing text-based interventions for social change, but such interventions rely on expert knowledge that is scarce in developing contexts. Expert knowledge can be scaled using chatbots, yet more research is needed to support users who need context-specific support, local-language interventions, or who may not have regular internet connectivity. To support the design of chatbot-based interventions in low-resource contexts, we built DIA, a chatbot architecture that scales expert knowledge and supports localization. DIA is a human-chatbot hybrid system that organically learns topic-specific knowledge and local language from user interactions. We built a preliminary version on WhatsApp and deployed it to mentor 38 teachers in rural Côte d’Ivoire. The deployment shows that DIA can help build datasets of topic- and language-specific dialogues, reveal users’ online smartphone usage through chat logs, and collect survey data through conversational interaction.
Citation
Cite this paper
The complete BibTeX record is visible below for readers and scholarly agents.
@inproceedings{cannanure2020dia,
title = {DIA: A Human AI Hybrid Conversational Assistant for Developing Contexts},
author = {Cannanure, Vikram Kamath and Brown, Timothy X. and Ogan, Amy},
booktitle = {Proceedings of the 2020 International Conference on Information and Communication Technologies and Development},
articleno = {24},
year = {2020},
doi = {10.1145/3392561.3397577}
}