Sentiment classification for employees reviews using regression vector- stochastic gradient descent classifier...

Area of Science:

• Natural Language Processing
• Machine Learning
• Organizational Psychology

Background:

• Employee satisfaction is vital for organizational productivity and goal achievement.
• Analyzing employee feedback through sentiment analysis can reveal satisfaction levels.
• Traditional methods may not fully capture the nuances of employee sentiment in reviews.

Purpose of the Study:

• To classify employee review sentiments for major global companies.
• To propose and evaluate a novel hybrid sentiment classification model.
• To compare the performance of the proposed model against other machine learning techniques.

Main Methods:

• Utilized TextBlob for initial sentiment extraction and labeling.
• Developed a hybrid Regression Vector-Stochastic Gradient Descent Classifier (RV-SGDC) using logistic regression, support vector machines, and stochastic gradient descent with majority voting.
• Employed term frequency-inverse document frequency (TF-IDF), bag of words, and global vectors for feature extraction.
• Evaluated models using accuracy, precision, recall, and F1 score.

Main Results:

• The proposed RV-SGDC model achieved a high accuracy score of 0.97.
• The hybrid architecture of RV-SGDC demonstrated superior performance compared to other tested machine learning models.
• Term frequency-inverse document frequency (TF-IDF) proved to be an effective feature extraction technique for this task.

Conclusions:

• The RV-SGDC model is a highly effective tool for classifying employee review sentiments.
• Hybrid models combining multiple machine learning algorithms can enhance sentiment analysis accuracy.
• Accurate sentiment analysis of employee feedback can inform organizational policies and improve workplace environments.