The Brazilian scientific community has available a curricular information system called Lattes. The analysis in the area of scientific production is rapidly evolving as we have an increase in the number of publications combined with increasingly interdisciplinary sources of information. The Brazilian scientific community has available a database with curricular information called Lattes which contain information about academic and scientific achievements. There is a demand to carry out the search and retrieval of information from the Lattes curriculum in order to reduce the difficulties in exploring bibliometric data and to aid in the analysis of this information. This article uses textual data mining techniques, multidimensional projection techniques and visualization based on analysis of bibliometric data extracted from the Lattes platform. The results obtained were clusters of research groups that are related to each other and clusters of research groups that work with very different themes. These results can become a tool for visually exploring bibliometric data from the Lattes platform.