This article presents one of the results of the research project that aimed to detect patterns of violent and georeferenced events, based on the information stored in the Crime Observatory of the Municipality of Pasto (Colombia) with data mining techniques. To georeference each event and process it geographically, a geocoder of urban addresses of the municipality of Pasto was built, under free software. This geocoder was integrated into a cartographic viewfinder, which by applying the grouping technique with the k-means algorithm, allowed to visualise and identify patterns in areas of the municipality where the different violent events occur. This information will make it easier for government and security agencies to make effective decisions related to citizen safety and the prevention of such events.