Implementación de un modelo de recomendación para la venta de flores aplicando técnicas de aprendizaje de máquina. un caso de estudio.

Abstract

It is well known that online stores such as Amazon, eBay, Walmart, Alibaba, BestBuy, Flipkart, Ikea and many more have exploited the advantage of recommendation algorithms to increase their sales and profitability. Therefore, when an e-commerce has a recommendation system that leads the visitor to view items of interest as well as related products according to their preferences, the probability of purchase increases. Additionally, that will allow the clients to only choose articles based on their searching and not be overhelmed by endless product lists, so that is how the web commerce sites have more options to achieve their sales. This means the recommendation system should identify the customer needs to adjust the items of interest that are shown. This is achieved when large amounts of data are processed using machine learning techniques to determine which elements are highly likely to be purchased. This project aims to build a model using the Company SAGAT INTERNATIONAL LLC database, and exploit data mining techniques to obtain the most optimal model in the predictions of elements of interest, evaluating the Apriori algorithms of association rules and K-means Clusters. The strategy of this project proposes the development and installation of the recommender model so that it interacts with the company’s e-commerce either within the store’s architecture or outside it through the consumption of a REST service.