mayureshd-18 / geospatial-crime-analysis-and-prediction

Crime and Criminal Analysis System integrating geospatial, temporal, and demographic analytics for predictive modeling of criminal activities. It employs machine learning for optimizing police resource allocation and incorporates real-time social media scraping for proactive crime detection.

Crime Analysis and Prediction

Overview

This project aims to provide comprehensive analysis and insights into crime patterns and trends using various techniques such as map-based analysis, month-based analysis, age-based analysis, crime prediction, crime sentiment analysis, and a model for resource allocation, crime prediction, sentiment analysis for social media comments for finding probable criminal activities.

System Architecture

Features

Map-Based Analysis

• Utilizes maps to visualize crime hotspots, aiding in understanding geographic patterns of criminal activities.
• Provides detailed views of crime distribution across different regions, enabling targeted interventions and resource allocation.

Month-Based Analysis

• Analyzes crime data based on months to identify seasonal trends and variations in criminal activities.
• Helps in understanding temporal patterns and planning preventive measures accordingly.

Age-Based Analysis

• Focuses on age demographics to study crime trends for women and sensitive age groups like the elderly and children.
• Provides insights into vulnerable age categories and potential areas for social interventions.

Crime Predictor

• Utilizes machine learning models to predict potential crime occurrences based on historical data and relevant factors.
• Aids law enforcement agencies in proactive planning and resource allocation.

Crime Sentiment Analysis

• Analyzes public sentiment related to crime through social media and other sources.
• Provides a sentiment score to gauge public perception and concerns regarding safety and security.

Resource Allocation Model

• A model for finding the nearest police station containing needed resources for the user-reported crime.
• Maximizes efficiency in resource utilization and response to varying crime scenarios.

Criminal prediction Model

• A model for predicting the most probable criminals based on the previous data for criminals including the time range, the locations of the crimes, and the crime type.

Tech stack

• Javascript: Integrating Google Maps API
• Python: Machine learning models
• Html/CSS: Website design
• Flask: Prediction model deployment
• Firebase: Realtime data storage and deployment
• Streamlit: Other models deployment
• Huggingface: Other models deployment
• AWS Amplify: Website deployment

Installation

1. Make Sure you have Python 3.x and Node.js installed in your system installed on your system.

2. Clone the repository:

   git clone https://github.com/CID123456/crime-analysis-project.git
   

3. Install dependencies in backend:

   cd backend
   pip install -r requirements.txt
   python app.py
   

4. Install live-server

   npm install live-server
   

5. Run in root directory (containing the index.html):

   live-server
   

Demo video

Demo video link - https://youtu.be/gC80tkqQN64