Halal Food Insight Dashboard

Sentiment Analysis & Theory of Planned Behavior (TPB) Factor Assessment

RESEARCH PROJECT: “HALAL FOOD PURCHASE DECISION PREDICTION FRAMEWORK FOR SOCIAL MEDIA POSTING USING MACHINE LEARNING”

What this dashboard does

The Halal Food Insight Dashboard uses AI (Natural Language Processing) to analyze public discussions about halal food across social media. It reveals overall sentiment and the TPB factors—attitudes, social influence, perceived control, and religious knowledge—that shape purchasing decisions. These insights help businesses, policymakers, and certification bodies respond strategically, address concerns, and strengthen trust in the halal ecosystem. You can explore trends, track sentiment shifts, and generate culturally and ethically aware recommendations.

How it works (at a glance)

1) Text Classification (Aspect-Based Extraction)

The system first categorizes each post to enable deeper analysis. It detects:

  • Opinions or experiences related to halal food
  • Religious views or knowledge
  • Questions, comparisons, or general feedback
  • Salient keywords
2) Sentiment Analysis

Each post is scored across multiple dimensions (trust, price, brand, compliance):

Positive: Confidence, trusted brands, fair pricing.
Negative: Doubts, high prices, distrust.
Neutral: Info seeking, general queries.
3) TPB Factor Mapping

Each post is mapped to relevant TPB constructs using advanced NLP:

Attitudes (Sikap)

Evaluative judgments (taste, cleanliness, quality).
Keywords: sedap, enak, bersih, kualiti, healthy.

Subjective Norms

Social influences (friends, family, viral trends).
Keywords: keluarga, netizen, viral, trend, "orang kata".

Perceived Behavioural Control

Perceived ease/difficulty and verification actions.
Keywords: susah, mudah, logo, scan, semak, verify.

Religious Knowledge

Knowledge of halal–haram principles and authorities.
Keywords: syariah, fatwa, JAKIM, sijil, gelatin.

4) Purchase Likelihood Prediction

Combining sentiment and TPB signals, the system estimates likelihood as: High, Medium, or Low.

Dashboard Details

TARGET USERS
  • Businesses/Brands: Optimize messaging/claims; track campaign impact.
  • Policymakers/Agencies: Identify gaps; target outreach; evaluate responses.
  • Certification Bodies: Monitor perceptions; spot recurring pain points.
TECHNICAL SPECS
  • Data Sources: Social media posts.
  • Languages: English, Malay, mixed-language (slang/colloquialism).
  • Outputs: Sentiment scores, TPB factors, Purchase Prediction.
Theory of Planned Behavior (TPB) Framework Depth
Attitudes

Reflect a positive perception of Halal food based on personal feelings and past experiences. Tied to religious values, health benefits, and food quality. These views shape how people express support or concern online.

Subjective Norms

Highlight the role of social influence (community, workplace, peer expectations). Reveals tension when social norms clash with personal beliefs. Reflects "what people think others expect of them" and collective concerns influence behavior.

Perceived Behavioral Control

Refers to how easy or hard individuals feel it is to follow a Halal diet. Includes personal confidence, motivation to comply despite practical barriers, and how digital tools (apps/reviews) support Halal decision-making.

Religious Knowledge

Indicates a deeper understanding of Halal in Islam, including ethical and environmental values. Encourages conscious food choices based on faith and drives more cautious sentiment on social platforms.

🎯 Why Use This Dashboard?

✓ Understand real-time public sentiment on halal food
✓ Identify key concerns and influencers
✓ Support certification bodies, businesses, and policymakers
✓ Promote ethical AI use in culturally sensitive topics