This research project aims at the design of an intelligent system for recognizing extortion text messages using natural language processing (PNL). The data on which we have worked is a sample of extortion text messages and not extortion text messages as well. The project is based on extortion messages, identifying during the process the mainly wrong uses of language while writing a message, which does not allow an accurate classification; due to this, we have used natural language processing to solve these problems and recognize correctly extortion messages. Finally, we have realized tests on the system using messages of our data base getting expected results