Proposal for modeling a publisher profile and implementation of a content recommendation system
Recommender Systems, machine learning techniques
This dissertation proposal aims to build a content recommendation system for a Brazilian publisher with a digital subscription club, where it offers its subscribers paid content in weekly editions. To develop the proposal, an analysis will be carried out on recommendation systems and unsupervised learning algorithms, in addition to the study for the domain of users who consume paid content in the digital environment by subscription clubs, considering mainly information on the browsing history of the customer on the company's website to determine their usage profile. In this study, we intend to provide the newspaper with knowledge about its subscriber base and a recommendation system that, if applied in production, can deliver suggested content filtered according to user preferences and generate action plans on how its content will be worked. for public engagement.