Currently, I am involved in the following projects:
Text Categorization with spaCy and sklearn: Developing a text categorization system using machine learning algorithms such as Logistic Regression, Random Forest, and Naive Bayes. At my company, our Customer Support department interacts with both our customers and Amazon support. At the end of each month, this department needs to prepare a report about the topics of the messages they've handled. Customer support agents assign descriptive keywords to the message topics. However, due to a lack of standardized procedures, there are numerous misspelled and inconsistently varied labels. Subsequently, someone manually filters the messages and prepares a report for customers.
Tasks to complete for this project include:
- Feature engineering (completed),
- Data exploration,
- Model benchmarking (completed),
- Data preproccesing (completed),
- Creation of a spaCy model (completed),
- Hyperparameter tuning using grid search for sklearn models.