AML — Anticipation Market Lab

AML (Anticipation Market Lab) is a web-based prediction market platform built around real-world events. Users predict event outcomes by staking the virtual currency AuraCoins—no real money is involved. The core of the application is a personalized feed of active Yes/No events with dynamic odds and voting deadlines.

The project was developed for the Principles of Software Engineering (SI3PSI) course at the University of Belgrade School of Electrical Engineering. This repository covers the complete project lifecycle: the initial idea, requirements specification, prototype, formal inspection, database design, implementation, UML models, and testing.

Team

This project was created collaboratively by a four-member team:

• Andrija Trnavčević
• Luka Pantović
• Vuk Bojović
• Mihailo Mandić

The work was not divided into isolated, member-owned modules. Every team member participated in every project phase and contributed to the analysis, documentation, design, implementation, and testing. Decisions and responsibility for the final solution were shared by the entire team.

Features

Guest

• Browse active events and current odds.
• Receive a registration prompt when attempting to vote.

Registered user

• Register, sign in, sign out, and reset a forgotten password.
• Select and update topics of interest.
• Browse a personalized event feed.
• Search and filter events by title and category.
• Vote Yes or No by staking AuraCoins.
• View odds that change dynamically based on total stakes.
• Check the wallet balance and transaction history.
• Review active and completed votes.
• View prediction accuracy and profit/loss statistics.

Arbitrator

• Apply for an arbitrator account by submitting a CV.
• Review event suggestions retrieved from Polymarket.
• Create and publish events manually.
• Close events and determine their final outcome.
• Trigger automatic payouts to winning users.

Administrator

• Approve or reject arbitrator applications.
• Ban and reactivate user accounts.
• Manually adjust AuraCoin balances.
• View aggregate platform statistics.
• Review an audit trail of important administrator and arbitrator actions.

Project phases and documentation

Phase	Repository contents
Initial idea and project proposal	AML.docx — problem statement, users, features, constraints, and priorities
Requirements and prototype	SSU/ — use-case specifications; Prototip/ — HTML screen prototypes
Formal inspection	FR/ — inspector logs, meeting report, and defect report
Database design	Baza/ — specification, MySQL Workbench model, SQL dump, and model image
Web application implementation	Django applications: accounts, events, betting, wallet, and core
UML modeling	UML/ — use-case, class, and sequence diagrams in PNG and PlantUML formats
Testing	tests/, tests inside the Django applications, and VAMP Katalon testiranje.krecorder

Technology stack

• Python 3.14.6
• Django 6.0.7
• MySQL and mysqlclient
• HTML, CSS, and JavaScript
• Selenium 4.46.0
• Polymarket API for real-world event suggestions
• PlantUML for UML models
• MySQL Workbench for database modeling
• Git for version control and team collaboration

Exact Python package versions are listed in requirements.txt.

Project structure

.
├── accounts/       # users, roles, and arbitrator applications
├── events/         # events, feed, odds, creation, and resolution
├── betting/        # votes and voting history
├── wallet/         # AuraCoin wallets and transactions
├── core/           # administration, audit logs, and statistics
├── templates/      # shared Django templates
├── static/         # CSS, JavaScript, images, and fonts
├── tests/          # functional, integration, and Selenium tests
├── Baza/           # database model and documentation
├── FR/             # formal inspection artifacts
├── Prototip/       # interactive HTML prototype
├── SSU/            # use-case specifications
├── UML/            # UML source files and exported images
├── AML.docx        # initial project proposal
├── manage.py
└── requirements.txt

Local setup

Prerequisites

• Python 3.14 (the project was developed with Python 3.14.6)
• MySQL 8
• Google Chrome for Selenium tests

1. Create a virtual environment and install dependencies

python -m venv .venv

Activate it in Windows PowerShell:

.venv\Scripts\Activate.ps1

Activate it on Linux or macOS:

source .venv/bin/activate

Install the required packages:

python -m pip install -r requirements.txt

2. Configure the database and environment variables

Create an empty MySQL database:

CREATE DATABASE AMLDB CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;

Copy .env.example to .env and enter your local database credentials:

SECRET_KEY=enter-a-local-secret-key
DEBUG=True

DB_NAME=AMLDB
DB_USER=your_mysql_user
DB_PASSWORD=your_mysql_password
DB_HOST=127.0.0.1
DB_PORT=3306

The .env file contains local secrets and must not be committed to GitHub.

3. Apply migrations and run the application

python manage.py migrate
python manage.py runserver

The application will be available at http://127.0.0.1:8000/.

The migrations create a development administrator account with the username admin and password admin. Change these credentials before deploying the application publicly.

4. Fetch event suggestions

To retrieve current event suggestions from the Polymarket API:

python manage.py fetch_suggested_events --limit 100

This command requires an internet connection. An arbitrator can review the retrieved suggestions, complete the required information, and publish a selected event.

Testing

Run the complete Django test suite with:

python manage.py test

The test suite covers models, controllers, access control, input validation, registration and authentication flows, the event feed, dynamic odds, voting, wallets, administration, and event creation and resolution. Selenium scenarios run Chrome in headless mode, so Chrome must be installed and available on the system.

An additional Katalon scenario is stored in VAMP Katalon testiranje.krecorder.

Notes

• AuraCoins are a virtual currency with no monetary value.
• Dynamic odds are a simplified prediction-market simulation, not an actuarial model for real-money betting.
• Baza/AMLDB_dump.sql is an artifact from the database-design phase. To run the current Django application, create an empty database and apply the Django migrations.
• User uploads, the local database, .env, virtual environments, and IDE settings are intentionally excluded from the repository.

Academic context

The project demonstrates the practical application of software engineering principles throughout a complete development process: requirements engineering, prototyping, formal inspection, database design, object-oriented and UML modeling, collaborative implementation with version control, and automated testing.