Supershop management Models: An optimised way to manage the supershop using hyperautomation and machine learning

Abstract

Customers are the heart and soul of supermarkets, and these stores often suffer losses due to mishandling of customer service. This study aims to examine how the satisfaction of customers can be maximized with the help of hyper-automation technologies in order to operate the stores successfully. Here, an intelligent voice bot is used to reduce response time for basic customer queries by providing real-time replies using NLP and for further complex queries, customers will be provided with contact info or forwarded to relevant authorities. In addition to that we predicted customer demands for future products by using multiple machine learning libraries like XGBoost, Linear Regression and Random forest with the help of daily sales data. This helps supermarkets to stock a perfect amount of products in their in ventory which will help them to avoid any kind of product shortage in any season and will help them to achieve customer satisfaction. Furthermore, optimized prod uct placement will be ensured with the use of data mining techniques like Apriori algorithm ,FP Growth algorithm and GSP algorithm. By doing this we are making it easy for the customers to find out their preferred product together in a single shelf and reducing customer hassle of iterating through the whole shop to find the products from their shopping list. As well as, to ensure a hassle-free transaction between consumer and seller, a system is proposed using Smart Contract System via Block chain which can make the process faster, and ensure transaction safety at the same time.For results, With an R-Squared score of 0.963, we discovered that the hybridization of linear-boost regression was the most suited for forecasting. The best outcomes for product placement were provided by FP Growth. For the pur poses of the chatbot, let’s say that for the two strings ”rfl nipple 3-6 month” and ”rfl nipple 3 to 6 month,” the spaCy llibrary and nltk’s bleu function both yield 90.8 and 66.21 percent similarity, respectively. Now, based on the %, you could assume that the spaCy library is operating more effectively, but this is untrue. SpaCy library functions. better in a big model where pre-trained word vectors are present, but not in a small model. However, the smart contract system successfully carried out all of the system’s algorithms and guaranteed transaction security as well as product safety. By using the Hyper Automation technology Super stores can ensure better customer service. The budget and implementation of such technologies combined into a system turned out to be both high and complicated. However, as time passes, the cost of these technologies will decrease rapidly and their usage will be further simplified.