The role of the coolhunter (trendsetter) has gained importance in the fashion world. This role is in charge of detecting when new trends are occurring. Social media, with its blogs, forums, social platforms, etc., has made the information about clothes and trends more accessible. The task of filtering the relevant from the irrelevant is very difficult for the coolhunter, due to the avalanche of data and information produced in these media. Therefore, there is a need for an automatic system for detecting trends. Trends, or clothing style recognition, is a challenging problem due to wide variation in clothing item appearance, and even the subjectivity associate with the style concept. In fact, clothes that people wear contain latent fashion concepts capturing styles. Fashion theorists have proposed that these concepts are shaped by design elements such as color, material, and silhouette. While a dress may be “classic” because of its color, material, or some combination thereof, it is not always clear how low-level elements translate to high-level styles. This paper presents the firsts steps towards a system that helps in the task of trend determination. The paper presents a tool which has learnt to classify a style in a picture. Specifically, it is able to determine whether the clothes in an image represent a classic outfit or a modern outfit. This system is a cognitive system in the sense that it has been improved with data from a test designed to capture what people think it is classic or modern. We also provide the results of some experiments done using supervised machine learning techniques.