Recommendation engine and digital marketing automation in e-commerce with machine learning

Electronic commerce (e-commerce) has become a pillar of the modern economy, characterized by increased competition and increasingly demanding consumers. In this context, digital marketing and recommendation systems are emerging as strategic tools to differentiate, optimize the customer experience, and improve business performance. This thesis examines the state of the art of recommendation systems and automated digital marketing in e-commerce. It explores different types of recommender systems, automated digital marketing techniques and their integration to improve customer engagement and business outcomes in the e-commerce industry. In addition, an approach combining association rules and collaborative filtering to improve recommendations by adding inference data to user profiles is proposed and their performance is scaled. According to the mathematical metrics RMSE and MAE, collaborative filtering outperforms the hybrid method used alone by a small margin. However, the hybrid system was observed to have more recommendation coverage and more recommendation diversity during the test.