Multi-objective Optimization for Order Assignment in Food Cart Evaluation System

Abstract

The growing trend of busy scheduling in everybody’s lifestyle has evolved the concept of the door-to-door delivery system. So, keeping that in mind and food being an essential commodity of life, we have seen that the food delivery industry has boomed in recent years. However, finding the best quality food at a minimum cost is a problem of interest. It is complicated for any customer to decide where to buy the best food at optimal cost if there is no such rating concept for the food. There is no concrete mechanism to help customers with the recommendation process. In this paper, we have designed a model that would help customers evaluate the various food carts that are serving in different locations and generalize the cost and overall rating based on service and food quality. It would basically work in the concept of minimizing the cost and maximizing the overall rating of the food cart. In order to get food cart evaluations, a multi-objective optimization problem is formulated, which includes mainly three multi-objective evolutionary optimization algorithms, namely NSGAII (Nondominated Sorting Genetic Algorithm II), SPEA II (Strength Pareto Evolutionary Algorithm 2) and IBEA (Indicator-Based Evolutionary Algorithm) for the generation of approximated Pareto solutions of our proposed model. The proposed work contributes to an analytics-based approach, mainly based on evolutionary algorithm solutions, to create a multi-objective food cart evaluation system. We used the Zomato dataset to validate the results for comparison and evaluation purposes.