Halal Assessment System
MCA + Spectral Clustering Experiment & Critical Ingredient Detection
8-Cluster Food Taxonomy Results (MCA + Spectral Clustering)
Research Methodology
Halal Industry Science
This application demonstrates objective halal evaluation based on ingredient composition recipes.
Why MCA? MCA (Multiple Correspondence Analysis) projects binary ingredient categorical data into a low-dimensional continuous space with Chi-square weighting, eliminating the effect of common ingredients (e.g., water or salt) that dominate similarity.
Why Spectral Clustering? Spectral Clustering partitions food product regions using eigendecomposition on an RBF proximity kernel in MCA continuous space, producing regular and stable groupings.
Why Ingredient Audit? Since 100% of halal certificates on complete-ingredient products in the ontology are empty, this analysis injects critical additive rules for independent self-audit.
Halal Product Proximity Topology Map
Click a point on the plot to inspect ingredient compositionSelect a Product
Click a point on the UMAP map
Enter Ingredient Composition
Enter the ingredient list of the commercial food product you want to audit. Separate ingredients using commas ( , ).
⚠️ IMPORTANT NOTICE: Ingredients MUST be entered in Professional English (e.g., "wheat flour", "beef", "pork fat") because the semantic rules and database taxonomy strictly operate in English.
Audit Report
Ingredient-by-Ingredient Status
No data scanned yet. Please enter ingredient composition on the left.
Halal Product Database Explorer
| Product Name | Manufacturer | Cluster | Certificate | Objective Halal Status | Composition |
|---|