In recent years, the number of biomedical publications freely available in the literature has grown enormously, resulting in a rich source of untapped new knowledge. However, most biomedical data is buried in the form of unstructured text, and their exploitation requires expert knowledge and time-consuming manual curation of published articles. Hence the development of novel methodologies that can automatically analyze textual sources, extract facts and knowledge, and produce summarized representations that capture the most relevant information in a timely fashion. INtERAcT represents a novel approach to infer interactions between molecular entities extracted from the literature using an unsupervised procedure that leverages recent developments in automatic text mining and analysis. INtERAcT implements a new metric that acts on the vector space of word representations to estimate an interaction score between two molecules.