Politicians, journalists and public figures are often exposed to hate speech in various digital environments, especially in different discussion forums and in comment fields. Detecting hate automatically is a difficult task since hate can be expressed in many different ways. In this paper we have developed a method to measure hate directed at politicians using a combination of natural language processing and automated reasoning Our method is adapted to work on Swedish, although in general it is language independent. We have tested our method in a study where we analyze hate directed at six Swedish politicians. The results shows that our method has a fairly high precision but a low recall compared to a manual assessment.