We currently have 3 databases of incident data and we are attempting to merge it all into one dataset, but the challenge is that some incidents may be reported in 1, 2, or all 3 systems. Looking for a model that can compare incident data and help determine if there is another similar incident so we don’t double count things. Right now we need to look at dates, times, names, locations, descriptions, etc. to determine this manually. Finally, due to data sensitivity, we can not have the data going external and being used as training data for models.
Thanks.