This work presents the design of an electronic document recommendation system for the Octavio Arizmendi Posada Library (BOAP) at the University of La Sabana, based on bibliographic text mining techniques. The system utilizes the Scopus API to retrieve the abstracts of documents consulted by a user, whose DOIs are recorded in the query logs, serving as a source for applying algorithms such as TF-IDF, Cosine Similarity, and K-Means. The main objective is to provide accurate and relevant recommendations tailored to the research interests and needs of its users. The information flow begins with the retrieval of data from Scopus, where abstracts of consulted documents are extracted. Subsequently, the TF-IDF algorithm is applied for feature extraction and the measurement of term importance in the documents. Then, the K-means algorithm is used to cluster the documents into thematic categories and construct a search equation that will be used to perform a new search within Scopus.