Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/PuzaTech/Fugue
A research package for topic modeling
https://github.com/PuzaTech/Fugue
latent-dirichlet-allocation topic-modeling topic-models
Last synced: about 1 month ago
JSON representation
A research package for topic modeling
- Host: GitHub
- URL: https://github.com/PuzaTech/Fugue
- Owner: PuzaTech
- License: apache-2.0
- Created: 2016-02-28T04:49:33.000Z (almost 9 years ago)
- Default Branch: master
- Last Pushed: 2020-11-14T18:14:52.000Z (about 4 years ago)
- Last Synced: 2024-08-03T18:20:57.839Z (4 months ago)
- Topics: latent-dirichlet-allocation, topic-modeling, topic-models
- Language: Java
- Homepage: http://puzatech.github.io/Fugue/
- Size: 5.15 MB
- Stars: 5
- Watchers: 3
- Forks: 0
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-topic-models - Fugue - Java implementation of collapsed Gibbs sampling with slice sampling for hyper-parameter optimization (Models / Latent Dirichlet Allocation (LDA) [:page_facing_up:](https://www.jmlr.org/papers/volume3/blei03a/blei03a.pdf))
README
# Fugue
[![Build Status](https://travis-ci.org/PuzaTech/Fugue.svg?branch=master)](https://travis-ci.org/PuzaTech/Fugue)
[![codecov.io](https://codecov.io/github/PuzaTech/Fugue/coverage.svg?branch=master)](https://codecov.io/github/PuzaTech/Fugue?branch=master)
[![Known Vulnerabilities](https://snyk.io/test/github/puzatech/fugue/badge.svg?targetFile=build.gradle)](https://snyk.io/test/github/puzatech/fugue?targetFile=build.gradle)
[![Code Climate](https://codeclimate.com/github/PuzaTech/Fugue/badges/gpa.svg)](https://codeclimate.com/github/PuzaTech/Fugue)
[![Codacy Badge](https://api.codacy.com/project/badge/grade/79ec17fe542e4f6792a522c7a9c374b4)](https://www.codacy.com/app/hongliangjie/Fugue)
[![GitHub license](http://dmlc.github.io/img/apache2.svg)](./LICENSE)## A Topic Modeling Package
Fugue is a research oriented topic modeling package for text mining and machine learning practitioners, designed for understanding models and algorithms. The philosophy of the package is to include mature models and algorithms that represent the state-of-the-art of research in topic modeling.### Models/Algorithms
* Latent Dirichlet allocation (LDA) with collapsed Gibbs sampling
* Using "Estimate theta" method for computing perplexity in test documents.
* Using multiple MCMC chains to average results.
* Slice sampling for hyper-parameter sampling
* Optimizain methods for hyper-parameter tuning### Roadmap
For 1.0, See [Detailed Plan](https://github.com/PuzaTech/Fugue/wiki/Fugue-1.0-Roadmap).### Prerequisites
* Gradle
* JDK
* Python 3.7The design of the package is to limit the prerequesite packages and therefore you can easily experiment new models with minimum environments.
### Platforms
* Mac/Linux### How to Compile
```sh
$ python Fugue.py
```### Run AP example
```sh
$ python Fugue.py --task train
```### License
Licensed under an [Apache-2.0](./LICENSE) license.