🛡️ Anonymous Complaint Reporting System

🔒 Confidentiality Notice

This repository contains a publicly shareable frontend demonstration of the project. The available HTML file includes the frontend structure, embedded CSS styling, and a small amount of JavaScript code required for the demonstrated interface.

The backend source code, complete JavaScript implementation, AI/NLP implementation, DistilBERT model files, database configuration, backend packages/dependencies, API credentials, environment variables, and other project-specific implementation details are confidential and have not been included in this repository.

The README provides a high-level overview of the complete system's architecture, workflow, modules, and technologies used for portfolio and demonstration purposes. The publicly available file does not represent the complete application.

An AI-powered anonymous complaint reporting system designed to provide a secure and confidential platform for reporting sensitive incidents such as child abuse, domestic violence, sexual harassment, corruption, and misconduct without requiring users to reveal their personal identity.

The system combines Natural Language Processing (NLP), DistilBERT-based sentiment analysis, location services, complaint tracking, and automated email communication to make complaint reporting easier, faster, and more accessible.

👥 Contributors

Mugesh Kumar .K
Manikandan .AR
Nandha Kishore .R
Kavin Prasanth .M

📌 Project Overview

Many victims hesitate to report abuse or harassment because of fear, social stigma, intimidation, emotional trauma, or pressure from family members.

Traditional reporting methods often require victims to directly approach authorities or provide personal information. This project proposes an anonymous digital alternative where users can submit complaints through:

🤖 AI Chatbot
📝 Anonymous Complaint Form

The submitted complaint is processed by the Flask backend and analyzed using a DistilBERT NLP model. Based on the classification result, the system can provide emergency-support guidance, register the complaint, generate an anonymous reference ID, and forward complaint information to the appropriate authority.

🎯 Objectives

Provide a confidential and anonymous complaint reporting platform.
Allow victims to report incidents without directly approaching authorities.
Use NLP and AI to analyze complaint text.
Detect emotionally distressing or potentially urgent complaints.
Provide immediate emergency-support guidance when appropriate.
Automatically forward complaints to the concerned authority.
Generate a unique anonymous Reference ID for tracking.
Provide a simple interface suitable for users under stressful situations.

✨ Key Features

🤖 1. AI Chatbot Complaint Collection

Users can describe their incident naturally through a chatbot interface.

The chatbot:

Collects complaint information conversationally.
Accepts natural-language descriptions.
Sends complaint text to the Flask backend.
Uses NLP processing for analysis.
Performs DistilBERT-based sentiment classification.
Provides appropriate feedback based on the analysis.

📝 2. Anonymous Complaint Form

Users can directly submit complaints through a predefined form.

The form can include:

Incident category
Location
Complaint description
Evidence attachments
Other relevant incident details

No unnecessary personal identification is required for anonymous reporting.

🧠 3. DistilBERT-Based Text Analysis

The system uses DistilBERT, a lightweight Transformer-based NLP model derived from BERT.

The model analyzes complaint text and performs sentiment/emotional classification.

The resulting classification is used as one input to the system's severity/urgency decision process.

Note: Sentiment analysis should not be treated as a standalone determination of whether a situation is an emergency. Real-world deployment should combine model predictions with contextual/risk indicators and appropriate human review.

🚨 4. Emergency Support

When the system identifies indicators of a potentially critical situation, it can display relevant emergency-support information to the user.

This provides immediate guidance instead of requiring the user to navigate away from the platform.

📍 5. Location-Based Authority Routing

Google APIs are used for location-related functionality.

The system can use the submitted location to help determine the appropriate local authority for handling the complaint.

📧 6. Automated Email Communication

EmailJS is used to send complaint information to the designated authority.

This reduces the need for manual forwarding of complaints.

🗄️ 7. Complaint Database

MySQL is used to store complaint-related records.

The database can maintain information such as:

Complaint ID
Complaint category
Complaint description
Location
Classification result
Severity/priority
Complaint status
Submission information

Sensitive data should be protected using appropriate access controls and security practices in production.

🔎 8. Anonymous Complaint Tracking

After submitting a complaint, the user receives a unique anonymous Reference ID.

The Reference ID can be used to check the status of the complaint without requiring the user to create a conventional personal account.

🏗️ System Architecture

                    ┌─────────────────────┐
                    │        User         │
                    └──────────┬──────────┘
                               │
                     ┌─────────▼─────────┐
                     │   Web Interface   │
                     │ HTML/CSS/JavaScript│
                     └─────────┬─────────┘
                               │
                  ┌────────────▼────────────┐
                  │      Flask Backend      │
                  │         Python          │
                  └────────────┬────────────┘
                               │
                     ┌─────────▼─────────┐
                     │   NLP Processing  │
                     │     DistilBERT    │
                     └─────────┬─────────┘
                               │
                    ┌──────────▼──────────┐
                    │ Severity / Decision │
                    │      Processing      │
                    └───────┬───────┬──────┘
                            │       │
              ┌─────────────┘       └─────────────┐
              │                                   │
      ┌───────▼────────┐                  ┌───────▼────────┐
      │   MySQL DB     │                  │ Emergency      │
      │ Complaint Data │                  │ Support Guide  │
      └───────┬────────┘                  └────────────────┘
              │
      ┌───────▼─────────┐
      │ Complaint        │
      │ Tracking System  │
      └──────────────────┘

              External Services
              ─────────────────
              Google APIs → Location Services
              EmailJS     → Authority Notifications

🔄 Data Flow

Complaint Input
      │
      ├── Chatbot
      │
      └── Complaint Form
              │
              ▼
       Flask Backend
              │
              ▼
        NLP Processing
              │
              ▼
          DistilBERT
              │
              ▼
      Sentiment / Emotion
          Classification
              │
              ▼
      Severity / Decision
          ┌────┴────┐
          │         │
       Normal    Critical
          │         │
          ▼         ▼
      Register   Emergency
      Complaint   Guidance
          │
          ▼
       MySQL
          │
          ├── Reference ID
          │
          └── Status Tracking
          
          └──────► EmailJS
                       │
                       ▼
                   Authority

📦 System Modules

Module 1 — Template-Based Complaint Collection

Provides a structured complaint form where users can select an incident category, provide details, optionally attach evidence, and submit the complaint anonymously.

A unique Reference ID is generated after successful submission.

Module 2 — Chatbot-Based Complaint Collection

Provides a conversational interface for users who prefer explaining their incident naturally instead of completing a structured form.

Module 3 — DistilBERT Classification

Processes complaint text using a DistilBERT-based NLP model to classify its emotional/sentiment characteristics.

Module 4 — Complaint Decision & Tracking

Uses the analysis along with the application's decision logic to determine the appropriate workflow, register the complaint, generate a Reference ID, and support complaint-status tracking.

Module 5 — Authority Communication

Uses EmailJS and location-related services to help forward complaint information to the appropriate authority.

🛠️ Technology Stack

Category	Technology
Frontend	HTML, CSS, JavaScript
Backend	Python, Flask
AI / NLP	NLP, DistilBERT, Sentiment Analysis
Database	MySQL
Location Services	Google APIs
Email Communication	EmailJS
Development Environment	Visual Studio Code
Version Control	Git / GitHub

📁 Suggested Project Structure

anonymous-complaint-system/
│
├── app.py
│
├── requirements.txt
│
├── models/
│   └── distilbert/
│
├── templates/
│   ├── index.html
│   ├── chatbot.html
│   ├── complaint.html
│   └── tracking.html
│
├── static/
│   ├── css/
│   │   └── style.css
│   │
│   └── js/
│       ├── chatbot.js
│       ├── complaint.js
│       └── tracking.js
│
├── database/
│   └── schema.sql
│
├── uploads/
│
└── README.md

⚙️ Installation

1. Clone the Repository

git clone https://github.com/your-username/anonymous-complaint-system.git
cd anonymous-complaint-system

2. Create a Virtual Environment

python -m venv venv

3. Activate the Environment

Windows:

venv\Scripts\activate

Linux/macOS:

source venv/bin/activate

4. Install Dependencies

pip install -r requirements.txt

5. Configure Environment Variables

Create a .env file and configure the required credentials.

MYSQL_HOST=localhost
MYSQL_USER=your_username
MYSQL_PASSWORD=your_password
MYSQL_DATABASE=anonymous_complaints

EMAILJS_SERVICE_ID=your_service_id
EMAILJS_TEMPLATE_ID=your_template_id
EMAILJS_PUBLIC_KEY=your_public_key

GOOGLE_API_KEY=your_google_api_key

Never commit API keys, database passwords, or other secrets to GitHub.

6. Configure MySQL

Create the required database and execute the SQL schema.

CREATE DATABASE anonymous_complaints;

Then configure the database credentials in the application's environment configuration.

7. Run the Flask Application

python app.py

The application will be available locally through the Flask development server.

🧪 Example Workflow

Normal Complaint

User submits complaint
        ↓
Flask receives complaint
        ↓
NLP preprocessing
        ↓
DistilBERT classification
        ↓
Complaint registered
        ↓
Reference ID generated
        ↓
Authority notification
        ↓
User tracks complaint

Potentially Critical Complaint

User submits complaint
        ↓
Flask receives complaint
        ↓
NLP + DistilBERT analysis
        ↓
Risk/urgency assessment
        ↓
Emergency-support guidance
        ↓
Complaint processing
        ↓
Authority notification
        ↓
Reference ID generated

🔐 Privacy & Security

Privacy is a core design consideration of this project.

The system is designed to:

Avoid unnecessary collection of personally identifying information.
Provide anonymous complaint submission.
Use a Reference ID instead of requiring a conventional user account.
Protect database credentials through environment variables.
Restrict access to sensitive complaint information.
Avoid exposing confidential complaint details through the frontend.

For a real-world deployment, additional security measures would be required, including:

HTTPS/TLS
Encryption at rest
Secure authentication for authorities
Role-based access control
Secure file upload validation
Audit logging
Data retention policies
Protection against SQL injection and XSS
Secure API-key management
Privacy and legal compliance
Human review for high-risk cases

⚠️ Important Disclaimer

This project is an academic/research prototype and should not be considered a replacement for police, emergency, medical, legal, or professional support services.

AI-based sentiment or emotion classification can produce incorrect results, particularly with sarcasm, ambiguous language, multilingual text, or complex descriptions. Therefore, an AI prediction should not be the sole basis for determining whether a person is in immediate danger.

📚 Research

The project is based on research related to:

BERT
DistilBERT
Natural Language Processing
Sentiment Analysis
Anonymous Reporting Systems
AI-based Complaint Classification
Secure Web-Based Complaint Management

Key References

Devlin et al. (2018) — BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding.

Sanh et al. (2019) — DistilBERT: A Distilled Version of BERT, a smaller and faster Transformer model.

Additional literature was reviewed on anonymous reporting systems, NLP-based sentiment analysis, and secure complaint management.

📊 Future Enhancements

Multilingual complaint processing
Voice-based complaint submission
Advanced emotion classification
Explainable AI for complaint classification
Human-in-the-loop authority verification
Encrypted evidence storage
Secure authority dashboard
Real-time complaint notifications
Improved risk/urgency classification
Mobile application
Integration with officially authorized emergency and government services

👨‍💻 Project

Anonymous Complaint Reporting System Using DistilBERT-Based Emotional Analysis

Built as an AI/NLP academic and research project combining web development, machine learning, database management, and external APIs.

Technologies: Python • Flask • HTML • CSS • JavaScript • NLP • DistilBERT • MySQL • Google APIs • EmailJS • GitHub

Name		Name	Last commit message	Last commit date
README.md		README.md
index.html		index.html

Latest commit

History

Folders and files

Repository files navigation