From Clicks to Carts: Developing an Autonomous E-Grocery Shopping System

Our research initiative, the Autonomous E-Grocery Shopping System with Machine Learning, is centered on the creation of an autonomous grocery shopping experience. One of the main predicaments with online shopping is the lack of personalization and support that is typically offered during in-person shopping. In physical supermarkets, products that are often purchased together are grouped together, incentivizing consumers to buy more and consequently increasing sales. We have incorporated this concept into our e-grocery shopping system to develop a platform that recommends items to users that they may not have been aware they needed. Our research focuses on the crucial role of autonomous artificial intelligence, with an all-encompassing assessment of various state-of-the-art techniques employed in the development of autonomous AI. Through our work, we have created a robust recommendation model utilizing the Apriori Algorithm of association rule mining, and Collaborative Filtering with the Nearest Neighbor’s algorithm. We have identified four major use cases, which include recommending grocery items based on users’ past purchase history, purchase history of similar users, similar items in the users’ cart, and recommending the highest-rated grocery items. We have also created a customized dataset and supported our model using a web application. The average value of support, confidence and lift for all the association rules are 0.001580, 0.124178, and 4.220583 respectively.