GIKI Mess Feedback Analysis Using NLP — training an ML model for complaint topic prediction
The candidate built an NLP-driven machine learning model to analyze and categorize complaints from GIKI Mess feedback. The workflow included transforming text to numerical features using TF-IDF and applying sentiment analysis and topic modeling with LDA to identify complaint topics. These activities represent supervised/derived labeling of text into structured topic outputs for downstream analysis. • TF-IDF vectorization of feedback text • Sentiment analysis to derive emotional polarity signals • Topic modeling with LDA to infer complaint topics • Model training on ~180+ feedback entries