On the early detection of threats in the real world based on open-source information on the internet

Many threats in the real world can be related to activities of persons on the internet. Internet surveillance aims to detect threats in an early stage and to assist in finding suspects based on information from the web. However, the amount of data on the internet rapidly increases and it is time consuming to monitor many open sources. In this paper, we present a method to find abnormal behavior on the internet and give an early warning for threats. The system was tested on Twitter data. The results show that it can successfully analyze the content of messages and recognize abnormal changes in sentiment and threatening content