Rajesh Neupane

Rajesh Neupane

PhD student Dairy Science (Computer vision/Machine Learning)

NLP-Dairy: Sentiment and Strategic Insights

Overview

This project applies advanced Natural Language Processing (NLP) techniques to the dairy industry to bridge the gap between consumer perceptions and industrial strategy. By analyzing large-scale textual data from social media, industry reports, and consumer reviews, we generate actionable insights that help stakeholders navigate the evolving landscape of dairy production and marketing.

Key Projects

1. Sentiment Analysis for Whole Milk: The “Whole Milk Comeback”

  • Objective: To understand the shifting consumer sentiment and perceptions towards whole milk products in the 2024-2025 market.
  • Approach:
    • Data Collection: Scraped social media (X, Reddit, Instagram) and consumer review platforms to capture “social conversations.”
    • Modeling: Employed sentiment classification models (VADER, BERT) and facial emotion analysis to identify drivers of consumer satisfaction.
  • Key Findings:
    • Sensory Superiority: Consumers consistently rate whole milk higher for “mouthfeel thickness” and “creaminess,” correlating with positive emotional responses.
    • Health Narrative Shift: Sentiment has moved away from viewing milk fat as a risk, with a 12% increase in conversations associating whole milk with “healthy fats,” hydration, and protein.
    • Aesthetic Rebellion: Identified a growing trend among Gen Z choosing whole milk as an “aesthetic rebellion” against plant-based alternatives.
  • Outcome: Provided brands with data-backed messaging strategies focusing on “nutritional density” and “clean label purity.”

2. Organic Dairy Systems: Navigating the Changing Landscape

  • Objective: To provide strategic insights for the organic dairy sector as it transitions from “basic organic” to “functional and high-tech organic.”
  • Approach:
    • Trend Analysis: Analyzed over 500 industry reports and regulatory documents using NLP-driven topic modeling (LDA).
    • Strategic Mapping: Developed a framework for understanding the “Clean Label 2.0” movement, focusing on Regenerative Organic and Grass-Fed certifications.
  • Actionable Insights:
    • Functional Growth: Identified high-growth opportunities in A2 organic milk and high-protein organic cheese, which are outperforming standard organic milk.
    • Supply Chain Transparency: Recommended the use of NLP-integrated systems to digitize and audit organic certification documents, ensuring consumer trust.
    • Sustainability Focus: Highlighted the shift toward “Food as Medicine,” where consumers value Omega-3 levels and antioxidant content in pasture-raised organic dairy.

Technical Stack

  • Languages: Python
  • NLP Libraries: NLTK, Spacy, HuggingFace Transformers (BERT, RoBERTa)
  • Data Analysis: Pandas, Scikit-learn
  • Visualization: Matplotlib, Seaborn, WordCloud
  • Techniques: Sentiment Analysis, Topic Modeling (LDA), Emotion AI, Named Entity Recognition (NER)