Flipkart collaborates with top Indian institutes for real-life research

In an effort to tap the next 200 million users or Bharat consumers, Flipkart is deepening its collaborations with some of the top institutes in the country. The e-commerce giant is looking at real-life research in areas such as artificial intelligence, machine learning, automation and natural language processing to solve some of the most critical industry problems with the help of professors and students. These collaborations are helping Walmart-owned Flipkart towards building a platform that is inclusive and accessible for Bharat customers. Majority of these users come from Tier-2 and Tier-3 cities and towns.
 

Flipkart has formed a memorandum of understanding with IIT Patna, as a step to strengthen Flipkart’s academic collaboration which it has been inculcating for past 5 years now. The partnership with IIT Patna will help create industry-focused applied research in the areas of AI, natural language processing (NLP) and machine learning (ML).
 

“The aim behind this collaboration is to create industry-focused applied research which could help reach e-commerce to more consumers and sellers alike,” said Mayur Datar, Chief Data Scientist Flipkart. “With this MoU, we aim to establish deeper academia collaborations which could help students and the academia to leverage our data and platform knowledge to work on India specific e-commerce challenges, in addition to publishing research papers.”
 

As a part of the collaboration, IIT Patna will undertake a number of programmes. These include joint research activities, writing research papers, internships and mentorship opportunities. The academic collaboration is expected to bring real-world industry exposure to students and scholars of IIT Patna. It would also provide an opportunity for the faculty members to work closely with Flipkart on research projects.
 

Asif Ekbal, Associate Professor, Department of Computer Science and Engineering, said that Flipkart has churned out many India specific and India first innovations to solve for the adoption of e-commerce in India. He said the research is particularly aimed at developing robust machine translation techniques for translating the large number of user reviews written in English to the Indian vernacular languages. At the same time, this will ensure that the translation process should preserve the domain knowledge such as sentiment, emotion and gender traits. He said this is extremely crucial for the translation service providers (TSPs) that make use of machine translation (MT) in production.
 

“Through this initial collaboration, we expect to come up with at least two high-quality publications in the relevant (areas), and create a baseline translation workflow in the product review domains,” said Ekbal. “This academic collaboration is expected to bring real-world industry exposure to the students and scholars of IIT Patna, and at the same time will provide an opportunity to the faculty members to work closely with Flipkart on research projects.”
 

Flipkart has collaborated with institutes such as the Indian Institute of Science (IISc) and various IITs and IIM for various projects. In the past, the solutions developed in collaboration with academia have been implemented for fraud detection or advertising and were put into production. It has also partnered with foreign universities such as Carnegie Mellon University and the University of California, San Diego to develop innovations related to areas such as fashion recommendation and query understanding. The other areas include demand planning, product categorization, review helpfulness, supply chain management and next basket prediction.
 

A former top research scientist at Google, Datar of Flipkart said academic collaborations are a great way to help accelerate the translation of research into new projects and help drive ecosystem benefits which could lead to economic growth.