The dataset consists, on the one hand, of a CSV file that includes the following fields: • ID_Film: Unique identifier for the film (IMDb ID).• Title: Name of the film.• Year: Year of release of the film.• Genre: Main genres of the film (e.g. action, drama, comedy, etc.).• Rating: Average rating given by IMDb users.• Votes: Total number of votes received on IMDb.• URL: Link to the film's page on IMDb• Image_URL: Link to the film's cover on IMDbAnd, on the other hand, a folder with all the downloaded images, which are named according to the ‘ID_Film’ field in the CSV file, ensuring that the image names match the identifiers for each film. The data included in the dataset will focus exclusively on films released in the 1980s, i.e. between 1 January 1980 and 31 December 1989, that are not intended solely for adults. The reason for this filtering, in addition to obviously reducing the size of the final dataset and not overloading the server, is based on the fact that, at that time, film posters were one of the main attractions of these films. In other words, all the digital marketing that is done today was concentrated, practically, on film posters and trailers, which could generally only be seen in cinemas.For all these reasons, this is one of the best periods for conducting this type of study on the influence of posters on the success of films. Only 250 films have been included, which is clearly too low a number to use in a CNN-based study, as it is only a sample. The programme allows you to specify how many you want to download in case you want to generate a larger dataset in the future.