https://github.com/ysenarath/opinion-lab
A place where opinions are experimented!
https://github.com/ysenarath/opinion-lab
cnn deep-learning iest implicit-emotion lstm natural-language-processing nlp opinion-mining opinions sentiment-analysis wassa wassa-2018
Last synced: 12 days ago
JSON representation
A place where opinions are experimented!
- Host: GitHub
- URL: https://github.com/ysenarath/opinion-lab
- Owner: ysenarath
- License: apache-2.0
- Created: 2018-08-06T07:15:00.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-06-15T09:59:59.000Z (almost 6 years ago)
- Last Synced: 2025-04-12T22:54:33.864Z (12 days ago)
- Topics: cnn, deep-learning, iest, implicit-emotion, lstm, natural-language-processing, nlp, opinion-mining, opinions, sentiment-analysis, wassa, wassa-2018
- Language: Python
- Size: 24.4 KB
- Stars: 3
- Watchers: 4
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# [Opinion Lab](https://ysenarath.github.io/opinion-lab/)
The repository contains codebase for building state-of-the-art deep learning techniques for opinion mining.
Currenly following models have been created and tested:
* Models for IEST @ WASSA-2018 [FNN, CNN, LSTM, CNN-LSTM, LSTM-CNN ++]
## Prerequisites
1. Related resources for featureizers (only if you are using them) (ex: lexicons, word-embedding models)
2. Some python knowledge
## Setup the Lab
1. Install [`textkit-learn`](https://github.com/ysenarath/textkit-learn/releases/tag/v0.2)
2. Configure path to resources in [`oplab/config.py`](https://github.com/ysenarath/opinion-lab/blob/master/oplab/config.py)
3. Happy Experimenting!
## How to Start
* To train a model its simple as calling
`python oplab train -c [config_file] -m [model_path]`
* To evaluate just
`python oplab evaluate -m [model_path] -e [evaluation_metrics] -o [predictions_file]`
## Licensing
* Code is under [Apache License 2.0](https://github.com/ysenarath/opinion-lab/blob/master/LICENSE).
* Model (releases) are under [Creative Commons Attribution - Non Commercial 2.0 Generic](https://creativecommons.org/licenses/by-nc/2.0/uk/legalcode) license.