Bug Prediction Using Learning to Rank (L2R)

A Python tool using machine learning to rank repository files by bug likelihood. It integrates Learning-to-Rank techniques for efficient bug prediction and code quality enhancement.

---

Features

• Data Extraction:
  Retrieve commit data from GitHub using the GitHub API.
• Data Preprocessing:
  Comprehensive cleaning, normalization, and feature engineering on commit data.
• Learning-to-Rank Implementation:
  Utilizes LightGBM to rank source code files based on bug likelihood.
• Model Evaluation:
  Evaluated using metrics such as NDCG score, precision, recall, accuracy, and F1 score.

---

Research Artifacts

For an in-depth look at the methodology, experimental evaluation, and outcomes, please refer to the following documents:

• Research Paper: Ranking Source Code for Bug Prediction: An L2R Approach on An Open Source Repository
• Presentation: Bug Prediction Using Learning-to-Rank

---

Getting Started

Prerequisites

• Python 3.x
• Pandas
• LightGBM
• Matplotlib
• Seaborn
• Scikit-learn

Installing

1. Clone the Repository:

git clone https://github.com/gillemta/bug-likelihood-ranker.git

2. Navigate to the Project Directory and Install Dependencies:

cd bug-likelihood-ranker
pip install -r requirements.txt

---

Usage

1. Insert Your GitHub API Bearer Token: In the main script, set your bearer token:

GITHUB_TOKEN = "<insert-token-here>"

2. Configure Data Retrieval: Modify the max_pages variable to specify the number of pages of commit data to pull (default is 50):

max_pages = 50

3. Run the Main Program:

python main.py

Bearer Token Setup

1. Log in to GitHub: Sign in to your GitHub account.
2. Access Token Settings: Click your profile picture in the top-right corner, then select Settings.
3. Developer Settings: Scroll down and click on Developer Settings.
4. Personal Access Tokens: Select Personal access tokens.
5. Generate a New Token: Click Generate new token, provide a descriptive name (e.g., "Bug Likelihood Ranker Token"), and select the required scopes (typically the repo scope).
6. Generate and Copy the Token: Click Generate token and immediately copy the token, as it will not be shown again.
7. Insert the Token: Paste the token into your program where indicated.

Troubleshooting

If you receive an error such as:

Error 401: {"message":"Bad credentials","documentation_url":"https://docs.github.com/rest"}

Double-check your token and ensure it’s correctly inserted. For more detailed guidance, refer to GitHub's documentation on personal access tokens.

---

Screenshots / Demo

Below is an example of the file that will be output after the program is ran