Question Classification System Using Naïve Bayes (NLP Project)

This project focuses on building a Natural Language Processing (NLP) model to automatically classify questions into predefined categories using Machine Learning techniques.

The model was trained on the TREC Question Classification Dataset, which includes multiple question types such as:

Human

Location

Number

Description

Entity

Abbreviation

The workflow of the project includes:

Loading and exploring the dataset using pandas

Performing text preprocessing techniques:

converting text to lowercase

removing punctuation

removing stopwords

tokenization

Converting text into numerical features using TF-IDF / Bag of Words

Training a Naïve Bayes classification model

Evaluating model performance and prediction accuracy

This project demonstrates practical implementation of NLP pipelines for automatic question classification systems used in:

Chatbots

Question Answering Systems

Search Engines

Virtual Assistants

Customer Support Automation

Technologies used:

Python

pandas

scikit-learn

NLP preprocessing techniques

TF-IDF / Bag of Words