2020_04_27_dataset.csv: 30633 tweets:
- 9959 Positive
- 6123 Negative
2020_05_24_dataset.csv: 53366 tweets:
- 19534 Neutral
- 17483 Positive
- 12376 Negative
- 3972 Mixed
- Server Go to server folder and run:
docker-compose up
- Client Go to client folder and run:
npm start
Go to model folder and follow the next steps:
- Setup
- Install virtualenv
pip install virtualenv
- Create new virtual env
virtualenv .
and activate itsource bin/activate
- Install dependencies
pip install emoji
- Run model/scraper.py
# Example
python3 model/scraper.py
- Run model/preprocess.py ${TODAY_DATE}
# Example
python3 model/preprocess.py 2020_04_27
- Send tweets_parsed to AWS Comprehend, download the output and rename it as
aws_output
, finally move it to model/data/${TODAY_DATE} folder - Run model/generate_dataset.py ${TODAY_DATE}
# Example
python3 model/generate_dataset.py 2020_04_27
- Create a folder called Sentime in Google Drive and open sentime.ipynb in Colab
- Upload the generated datasets to
sentime/${TODAY_DATE}/train/data.csv
andsentime/${TODAY_DATE}/test/data.csv
- Go to http://localhost:5431
- Login with email:
admin@test.com
and password:root_12345
- Create a new server
- Fill the form as follows:
- Use
db
instead oflocalhost
in the address field - User:
sentime_user
- Password:
root_12345
Go to http://localhost:5000/search/status
Go to http://localhost:5000/search/test?text=Juan%20est%C3%A1%20tirando%20c%C3%B3digo%20con%20cule%20de%20felicidad
Go to http://localhost:5000/users/signup/4f3a0ca08e906