CRIME ANALYSIS AND PREDICTION BASED ON ONLINE MEDIA NEWS USING DEEP LEARNING ALGORITHM

Theft, murder, assault, violence, rape, obscenity, kidnapping, vandalism, arson, incarceration, narcotics, fraud, corruption, embezzlement, and public order are all forms of crimes that can induce psychiatric illnesses in anybody. These are all natural phenomena and events that can have an impact on social circumstances and human existence. As a result, criminal activities performed by law enforcement agents must be prevented, and perpetrators must be apprehended and punished in accordance with their crimes.
The goal of this project is to use supervised learning to do analysis and prediction in order to monitor and detect all probable crimes in Indonesia. The neural network method is provided and a deep learning algorithm approach is used in this study to display crime data in order to obtain optimum crime data accuracy. Using this method, we can extract previously unknown and valuable information from unstructured data. Data processing activities and approaches include convolutional neural networks and recurrent neural networks.

Keywords: supervised learning, deep learning algorithm, convolutional neural network and recurrent neural network, neural network.