https://github.com/codito/obliviate

A collection of algorithms to model memory and forgetfulness
https://github.com/codito/obliviate

bayesian-statistics memory quiz retention spaced-repetition statistical-methods

Last synced: 13 days ago
JSON representation

A collection of algorithms to model memory and forgetfulness

• Host: GitHub
• URL: https://github.com/codito/obliviate
• Owner: codito
• License: mit
• Created: 2021-05-10T05:32:12.000Z (over 3 years ago)
• Default Branch: master
• Last Pushed: 2021-05-14T12:15:54.000Z (over 3 years ago)
• Last Synced: 2024-11-10T09:20:38.458Z (2 months ago)
• Topics: bayesian-statistics, memory, quiz, retention, spaced-repetition, statistical-methods
• Language: C#
• Homepage:
• Size: 36.1 KB
• Stars: 0
• Watchers: 4
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Obliviate

A collection of algorithms to model memory and retention of facts.

[![Build Status](https://github.com/spekt/testlogger/workflows/.NET/badge.svg)](https://github.com/spekt/testlogger/actions?query=workflow%3A.NET)

[![NuGet](https://img.shields.io/nuget/v/Obliviate.svg)](https://www.nuget.org/packages/Obliviate/)

## Usage

Install the nuget package in your project with `dotnet add package obliviate`.

### Ebisu

Ebisu provides a simple model that must be attached with each _fact_ the user is

trying to memorise. See the notes on [EbisuModel][] on choosing the parameters.

A learning/quizzing app will need to store the model, schedule reviews and keep

it fresh with observations from each review session. Ebisu provides two primary

APIs for these tasks. First, [PredictRecall][] attempts to find recall

probability of the existing model at a given time. E.g. _will I remember this

fact after X time units from the last review?_ Second, assume we reviewed the

fact `n` times with `k` successful reviews after `t` time units from last

review. [UpdateRecall][] updates the previous model with these additional

observations.

Ebisu provides fantastic documentation [here][ebisu]. We highly recommend a read

if you're planning to use the algorithm.

[ebisumodel]: https://github.com/codito/obliviate/blob/master/src/Obliviate/Ebisu/EbisuModel.cs

[predictrecall]: https://github.com/codito/obliviate/blob/54e74e55fd27bd4681c94bef8c60acd5f90aaabd/src/Obliviate/Ebisu/EbisuModelExtensions.cs#L29

[updaterecall]: https://github.com/codito/obliviate/blob/54e74e55fd27bd4681c94bef8c60acd5f90aaabd/src/Obliviate/Ebisu/EbisuModelExtensions.cs#L68

[ebisu]: https://fasiha.github.io/ebisu/

## Algorithms

- [x] Ebisu: https://fasiha.github.io/ebisu/ v2.0.0 (Public domain)

  - [ ] Ebisu v2.1.0 support with soft binary quizzes and half life rescale

- [ ] Memorize: https://github.com/Networks-Learning/memorize (MIT)

- [ ] Duolingo Halflife: https://github.com/duolingo/halflife-regression (MIT)

- [ ] SM-2 and related family of algorithms

We plan to support these algorithms along with benchmarks in future. Contributions

and suggestions are most welcome!

## License

MIT