Skip to content

SSahas/Sentiment-analysis-airline

Repository files navigation

Sentiment_analysis_airline

Using object oriented programming for training the model. Different scikit models are trained with one command.

pipeline

pipeline

  • This repository contains sentiment classification model , which takes text as input and outputs the whether the statement is positive or negative comment.

  • The app.py is main file of webapp deployed on huggingface, uses streamlit framework as ui.

  • Try the app here - https://huggingface.co/spaces/SSahas/sentiment_classifier_airline

  • The fast_api_swagger.py is code to create restful api using fastapi and uvicorn as server, swagger documentation is integrated with it, takes statement as input and outputs the prediction in the form of json.

Results

mean_fit_time std_fit_time mean_score_time std_score_time params mean_test_score std_test_score model_name
75.238871 7.649687 1.116081 0.396170 {'alpha': 0.9} 0.887878 0.010712 Ridge_classifier
32.970682 19.089618 0.537239 0.069481 {'alpha': 1.0} 0.887532 0.011108 Ridge_classifier
59.446653 4.041303 0.925700 0.067038 {'alpha': 0.8} 0.887098 0.010756 Ridge_classifier
4.578112 0.228420 0.208683 0.015610 {'max_depth': None, 'n_estimators': 10} 0.842732 0.049028 Randomforestclassifier
33.682326 6.913221 0.231637 0.023633 {'alpha': 9.5e-05} 0.838400 0.021505 SGDClassifier
56.504217 3.619727 0.524913 0.152613 {'max_depth': None, 'n_estimators': 150} 0.833893 0.096404 Randomforestclassifier
1.457530 0.724146 0.263000 0.058040 {'alpha': 0.5} 0.830604 0.016622 MultinomialNB
40.424593 12.100794 0.216021 0.011447 {'alpha': 8e-05} 0.830603 0.014611 SGDClassifier

About

Sentiment analysis classifier with scikit learn and NLP. Training models with OOPs(object oriented programming) and Gridsearchcv

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published