Abstract :

The Rapid Growth Of Conversational Artificial Intelligence (AI) Is Transforming Customer Support Systems, With The Global Market Expected To Reach USD 32 Billion By 2030 And Chatbots Projected To Handle A Majority Of Customer Interactions. Despite This Advancement, Manual Intent Annotation And Query Classification Remain Labor-intensive And Inconsistent, Posing Challenges To Scalability And Efficiency. This Study Proposes An Advanced Natural Language Processing (NLP) Framework Built On A Customer Support Bitext Dataset Annotated With Multiple Intents And Categories. The Approach Begins With Systematic NLP Preprocessing And Exploratory Data Analysis (EDA), Including Text Normalization, Tokenization, And Data Distribution Analysis To Prepare The Dataset For Modeling. For Feature Extraction, The Framework Utilizes A Miniature Language Model (MiniLM), A Lightweight Yet Context-aware Language Model Capable Of Capturing Semantic Relationships Effectively. To Address Class Imbalance, The Synthetic Minority Oversampling Technique (SMOTE) Is Applied To Generate Synthetic Samples For Underrepresented Classes, Ensuring A More Balanced Dataset. In Contrast To Machine Learning (ML) Methods Such As Decision Tree Classifier (DTC), K-Nearest Neighbors (KNN), And Naïve Bayes Classifier (NBC), The Proposed Model Incorporates Deep Neural Network (DNN)-based Feature Selection Combined With KNN Classification To Enhance Predictive Performance. The System Is Designed To Perform Bivariate Prediction By Identifying Both Intent And Category, Thereby Improving Contextual Understanding Of Customer Queries. The Trained Model Is Integrated Into A Chatbot Interface To Support Real-time Intent Detection And Automated Response Generation, Resulting In Improved Classification Accuracy And Reduced Inconsistencies In Annotation.

Published:

24-4-1-2026

Issue:

Vol. 26 No. 4-1 (2026)

Page Nos:

857-867

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite