/Employee-Sentiment-Tracker

Employee Sentiment Track is a project aimed at understanding organizational climate through the analysis of employee opinions. By leveraging sentiment analysis and topic modeling techniques, this project provides valuable insights into the workplace environment.

Primary LanguageCMIT LicenseMIT

Employee-Opinion-Tracker

Python version Latest version GitHub forks GitHub Stars License

Employee Opinion Tracker is a comprehensive aimed at understanding organizational climate through the analysis of employee opinions. By leveraging sentiment analysis and topic modeling techniques, this project provides valuable insights into the workplace environment. The results are presented through an interactive Streamlit dashboard, making it easy to visualize and interpret the data.

To use streamlit_app_openai_chatbot.py you need to insert your OPENAI_API_KEY in ./.streamlit/secrets.toml

Tree Structure

├── .conda
├── .streamlit
├── __pycache_
├── data_preprocessing
│   └── Text Cleaning & Normalization_Ali.ipynb
├── dataset
│   ├── Capgemini_Employee_Reviews_from_AmbitionBox.csv
│   └── data_employee_topic.csv
├── eda
│   └── EDA_Empl_Opinion_Tracker_r2.ipynb
├── sentiment_analysis
│   ├── preprocessing_capegini.xlsx
│   ├── sentiment analysis.ipynb
│   └── sentiment_evaluation_metrics.xlsx
├── style
│   └── style.css
├── topic_modeling
│   ├── Capgemini_Employee_Reviews_Topics.csv
│   ├── final_xgboost_sg_tuned.joblib
│   ├── topic_modeling_capgemini_notebook.ipynb
│   └── word2vec_sg.bin
├── user_employee_feedbacks
│   ├── user_employee_feedbacks.json
│   └── user_feedback.csv
├── .gitignore
├── LICENSE
├── README.md
├── SPARK - Employee Opinion Tracker.pdf
├── bert_sentiment.py
├── data_preprocessing.py
├── inference.py
├── logs.log
├── requirements.txt
├── Result_1.jpeg
├── Result_2.jpeg
├── streamlit_app.py
└── streamlit_app_openai_chatbot.py

Running the App

  • Don't forget to install all the required library

Running the streamlit_app.py:

streamlit run streamlit_app.py

Deployment

Deployment currently in local using Streamlit only. I have included two screenshoots how the Streamlit Dashboard looks like:

result_1 result_2

Detail Presentation

To see the detail Background & Problem statement, Objective & Scope, Workflow, Data Source, Model Evaluation & Performance, and Conclusion please see SPARK - Employee Opinion Tracker.pdf

License

Distributed under the terms of the MIT license, "Employee-Sentiment-Tracker" is free and open source software.

Issues

If you encounter any problems, please file an issue along with a detailed description.