https://github.com/fcakyon/glassdoor-review-textgenrnn
Train char-rnn with Glassdoor reviews and generate sentences
https://github.com/fcakyon/glassdoor-review-textgenrnn
aselsan char-rnn glassdoor nlp python reviews tai tensorflow text textgeneration textgenrnn turkcell turkish-airlines
Last synced: about 2 months ago
JSON representation
Train char-rnn with Glassdoor reviews and generate sentences
- Host: GitHub
- URL: https://github.com/fcakyon/glassdoor-review-textgenrnn
- Owner: fcakyon
- License: mit
- Created: 2021-01-14T20:04:35.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2021-01-14T20:11:35.000Z (over 4 years ago)
- Last Synced: 2025-02-08T09:47:24.166Z (3 months ago)
- Topics: aselsan, char-rnn, glassdoor, nlp, python, reviews, tai, tensorflow, text, textgeneration, textgenrnn, turkcell, turkish-airlines
- Language: Python
- Homepage:
- Size: 6.53 MB
- Stars: 1
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Introduction
This repo lets you train [char-rnn](http://karpathy.github.io/2015/05/21/rnn-effectiveness/) with Glassdoor reviews of [Aselsan](https://www.glassdoor.com/Reviews/Aselsan-Reviews-E41213.htm), [TAI](https://www.glassdoor.com/Reviews/Turkish-Aerospace-Industries-Reviews-E326234.htm), [Turkcell](https://www.glassdoor.com/Reviews/Turkcell-Reviews-E9709.htm) and [Turkish Airlines](https://www.glassdoor.com/Reviews/Turkish-Airlines-Reviews-E13316.htm)
Refer to [demo notebook]() for sample usage.
# Installation
First, make sure that you're using Python 3.
1. Clone or download this repository.
2. Run `pip install -r requirements.txt` inside this repo. Consider doing this inside of a Python virtual environment.
# Usage
Download Glassdoor reviews of any company using [glassdoor-review-scraper](https://github.com/MatthewChatham/glassdoor-review-scraper). Scraped reviews for Aselsan, TAI, Turkcell and Turkish Airlines are provided under `data` folder as `companyname_reviews.csv`.
1. Edit train/generator parameters for desired company, `nano aselsan_review_generator.py`
2. Train and generate reviews for desired company, `python aselsan_review_generator.py`
3. Generated texts will be saved to `textgenrnn_output_path`