This paper aims to develop a methodology for user classification and profiling by combining advanced machine, deep learning, and semantic analysis techniques. The main objective is to realize a pipeline capable of classifying a user from his or her online comments, assigning him or her to a specific type based on defined criteria. To achieve this goal, several basic steps will need to be followed. First, a thorough analysis of the state of the art was conducted to verify the existence of classifications already performed by experts in the field and to identify any established patterns. Next, an ontology was created to structure a solid basis for automatic classification. A further essential step was using natural language processing (NLP) techniques to extract, from a user’s comments, all relevant keywords that were processed in different ontologies depending on the context and type of analysis.