IIT Madras on board Google wagon driving AI for Social Good
Chennai-based technology institute IIT-Madras will collaborate in a project backed by search giant Google to develop AI-driven systems to predict engagement and churn in a phone-based information system meant for expectant mothers.
IIT Madras has tied up with Google for its “AI for Social Good” initiative and will handle two projects with accredited NGOs in the country, Balaraman Ravindran, who heads the Robert Bosch Centre for Data Science and AI at IIT Madras, told ET on Monday.
Another novel initiative that the institute is working on in collaboration with a team from AI4Bharat is for the NGO Storyweaver which is a digital repository that provides easy-to-use tools to read, create, adapt, and translate new stories for children in their mother tongue. The team from IIT Madras is being led by Mitesh Khapra and Pratyush Kumar and looks to build open-source input tools for underserved Indian languages to accelerate publishing of openly licensed content.
Google had announced that it would support six research projects led by organizations from India and across Asia with expertise in areas such as computer vision, natural language processing, and other deep learning techniques, in addition to funding and computational resources. The goal is to provide NGOs with technology that will enable them to address societal, humanitarian and environmental challenges using Artificial Intelligence.
The two Indian institutes that are participating in this project are IIT Madras and IIIT Delhi. IIT Madras is looking to use AI to help the NGO Armman to build predictive models that could prevent those who enrol for the healthcare programme from dropping out.
“One of the biggest challenges when you go outside metros in India is that there are a lot of superstitions (guiding pregnancy),” Dr Balaraman said. “People have no clear idea of what are the right processes, the correct progression, things to watch out for in early neo natal care and so on. A lot of information is avaialble online in English but even if people in rural areas are able to get online on their phones, they are not able to find information in a form that is accessible to them.”
The NGO works with expectant and mothers of newborns and has devised a model wherein those who enrol or the programme receive phone calls to improve targeted interventions and increase positive healthcare outcomes for mothers and their babies.
“The number of women who are enrolled in the programme is not a tiny number but we are trying to find out how many are truly engaged,” he said. “The predictive model will tell you things like how many calls will a woman answer before dropping out of the programme based on a variety of factors such as age, other demographic factors and their call behaviour like how long they listen to the first phone call.”
He said this could also help with making inferences that could help with understanding what kind of other interventions would be required. These could be having a human call when a woman is not engaging on a recorded call or even having an Asha worker visiting the family to check on the woman. However, he said doing this for everyone was neither scalable nor feasible.
In a blog post on the Google India page on February 13, Manish Gupta, Director, Google Research Team in India and Milind Tambe, Director AI for Social Good, Google Research Team, India spoke about how they believe AI can tackle some of the biggest issues that we face today.
“Healthcare, conservation, education, and disaster prediction are some of the most difficult challenges of our time. As computer scientists, it’s incredibly humbling and exciting to partner with the community towards making a positive impact for people in India and around the world,” they said.