Document Type


Publication Date



text mining, drug information, interactions, drug, disease


The Brazilian Ministry of Health and other related organizations are concerned with the issue of self-medication. Although these organizations warn about the risks of concomitantly using different drugs, they do not provide any tools to facilitate this process. ANVISA offers a collection of 6.076 medication guides in PDF file format. However, the information available in this guides are in an unstructured format. One of challenges of this work consisted in the automatic retrieval of information from ANVISAS’s medication guides. This paper presents a semiautomatic procedure that maps ANVISAS’s medication guides to DrugBank and SNOMEDCT. The medications, the diseases, the drugs, and their relations were structured and stored on a graph database using the Neo4j technology.


This paper is in Portuguese (Adicionando informações estruturadas ao Bulário Eletrônico da ANVISA)