The flow of information arises day by day through the Internet in a continuous way thanks to the constant interactions between users, these interactions present feelings that can be positive or negative. This helps social media content creators a lot to understand how useful what they do is for their followers, and if these are a large number, an analysis done by a single person is not enough. For this, it is necessary to use tools that operate with large amounts of data, such as BERT, which is a model that helps analyze sentiments and classify comments based on what one of them expresses. In this work, this model will be used for the classification of YouTube comments and the classification of videos on this same platform, evaluating these videos according to their content and helping viewers to choose the videos if they help them concerning what is expected. find searching. This work w...