URBAN CRIME INCIDENT TYPE CLASSIFICATION USING SPATIO-TEMPORAL AND CONTEXUAL ATTRIBUTES

Abstract

Urban areas are increasingly facing diverse crime incidents that significantly impact public safety and urban management. Effective classification and analysis of such incidents are crucial for improving law enforcement strategies and reducing crime rates. This project, titled “Urban Crime Incident Type Classification Using SpatioTemporal and Contextual Attributes,” aims to develop an intelligent system capable of automatically classifying crime types based on multiple influencing factors. The proposed approach utilizes historical crime data by integrating spatial attributes (such as location, latitude, longitude, and region), temporal attributes (including time, day, and month of occurrence), and contextual attributes (such as environmental and surrounding conditions). By combining these dimensions, the system identifies meaningful patterns and relationships associated with different types of crimes. Various machine learning algorithms are employed to train classification models that categorize crime incidents into types such as theft, robbery, assault, and burglary. Data preprocessing techniques—including data cleaning, feature extraction, and normalization—are applied to enhance data quality and improve model performance. The models are evaluated using standard metrics such as accuracy, precision, recall, and F1-score to determine the most effective approach.