softwareimpacts/simpac-2022-129

RatingsLib: A python library for rating methods with applications

RatingsLib is a Python library dedicated to rating/ranking systems implementation with applications in sports and other fields.

Installation

RatingsLib requires Python 3.8 or newer. More details about requirements can be found in requirements.txt. You can install ratingslib directly ::

 pip install git+https://github.com/ktalattinis/ratingslib

or by cloning the repository ::

 git clone https://github.com/ktalattinis/ratingslib
 cd ratingslib
 pip install .

Implementation

Rating/Ranking systems:

• WinLoss
• Colley
• Massey
• Keener
• Elo
• Offense - Defense
• GeM
• AccuRATE

Ranking Aggregation methods:

• Borda Count
• Average Rank

Rating Aggregation methods:

• Markov
• Perron
• Offense-Defense

Comparison metrics:

• Kendall's Tau

Applications & Examples:

• Sports (the main application of the library):

  • Soccer Teams rating
  • Soccer Teams ranking lists comparison
  • Hindsight and foresight prediction of the final outcome of soccer matches
  • Combining rating systems and machine learning methods to predict soccer matches outcome
  • Ranking NFL teams

• Other Applications & Examples:

  • Finance:
    • Examples from investment selection and portfolios rating and ranking.
  • Domain Market:
    • An illustrative example is provided and shows the ranking of domain names.
  • Movies:
    • Application on real-world dataset from MovieLens

Documentation

The documentation is available at: https://ktalattinis.github.io/ratingslib/