Module A: Sensory Attribute & Domain Concept Lexicon¶

Example User Question for Pihu:

"Compare the overall liking of SLHB5 and SLHB8 from the Signature Latte Cold V2 study. Did SLHB8 meet the benchmark, and what were the main aroma notes for it among millennials?"

This question is good because it involves:

Comparison: SLHB5 vs SLHB8
Specific Metric: Overall Liking
Specific Products & Study: SLHB5, SLHB8, Signature Latte Cold V2 study
Benchmark Evaluation: "Did SLHB8 meet the benchmark?"
Specific Attribute: Aroma Notes
Demographic Filter: Millennials
Implied Context: Sensory data from TagTaste

Now, let's start by rewriting Module A (Sensory Lexicon), focusing on the attributes relevant to this example question.

Module A: Sensory Attribute & Domain Concept Lexicon¶

Goal: To define all sensory terms and concepts Pihu needs to understand. This helps Pihu correctly interpret what users are asking about.

How this module helps answer our example question: "Compare the overall liking of SLHB5 and SLHB8 from the Signature Latte Cold V2 study. Did SLHB8 meet the benchmark, and what were the main aroma notes for it among millennials?"

Pihu needs to understand:

What "overall liking" means.
What "aroma notes" are.
How these are typically measured or described.

1) Text-Based Structure with Actual Examples (Relevant to the User Question)¶

Here, we'll show how entries for "Overall Product Liking" and a generic "Aroma Note" would look.

Attribute Entry Form: Overall Product Liking

Attribute ID: ATTR_LIKING_OVERALL_001
Canonical Name: Overall Product Liking
Category: Hedonic (Relates to pleasure/preference)
Status: Active
Definition: The overall degree to which a respondent likes or dislikes a product, considering all its sensory aspects (appearance, aroma, taste, texture, etc.) after evaluation.
Scope Notes: This is a holistic measure of preference. It's distinct from liking of individual attributes (e.g., "liking of aroma").
Perception & Evaluation:
- Mechanism: A subjective, integrated judgment based on the complete sensory experience and individual preference.
- Standard Evaluation Technique: Panelist rates their overall liking on a hedonic scale after interacting with/consuming the product.
- Evaluation Context: Typically asked at the end of a product evaluation sequence.
Synonyms & User Terms:
- Overall Liking
- Product Liking
- Liking Score
- Hedonic Score
- Preference Score
- How much they liked it
- Overall acceptance
Hierarchy:
- Parent Attribute ID: (blank)
- Child Attribute IDs: (blank)
Measurement Details:
- Typical Scale Types:
  - [X] Category Scale (e.g., 1-9 Hedonic)
  - [ ] Line Scale
- Typical Anchors (for a 9-point scale):
  - Scale Point: 1, Label: Dislike Extremely
  - Scale Point: 5, Label: Neither Like nor Dislike
  - Scale Point: 9, Label: Like Extremely
- Units: Score (e.g., out of 9)
Relationships:
- Related Statistical Concepts:
  - [X] Mean
  - [X] Median
  - [X] Top-2-Box Percentage
- Influenced By Factors:
  - All individual sensory attributes (Appearance, Aroma, Taste, Texture)
  - Product concept
  - Respondent's personal preferences and past experiences
- Influences Attributes:
  - ATTR_PURCHASE_INTENT_001 (Liking often influences purchase intent)
Examples:
- What is the overall liking for Sample A?
- Compare the overall product liking of SLHB5 and SLHB8.
Keywords for Search:
- liking
- preference
- hedonic
- acceptance
- overall score
Notes for Pihu: This is a key performance indicator. Often compared against benchmarks (see Module B) or other products.
Created By: sensory_team_lead
Created At: 2023-01-10T10:00:00Z
Last Modified By: sensory_analyst_01
Last Modified At: 2023-11-01T14:30:00Z

Attribute Entry Form: Aroma Note (Generic/CATA)

Attribute ID: ATTR_AROMA_NOTE_CATA_001
Canonical Name: Aroma Note (CATA)
Category: Olfactory
Status: Active
Definition: A specific, individual aromatic characteristic perceived orthonasally (by smelling before tasting) from a predefined list, typically identified using a Check-All-That-Apply (CATA) methodology.
Scope Notes: Refers to the presence/absence of specific aroma descriptors. Intensity might be captured separately or implied by frequency of mention. Examples: "Coffee-Roast," "Fruity," "Spicy."
Perception & Evaluation:
- Mechanism: Detection of specific volatile organic compounds (VOCs) associated with the descriptor.
- Standard Evaluation Technique: Panelists sniff the product and select all applicable aroma terms from a provided list.
- Evaluation Context: Orthonasal olfaction, before tasting.
Synonyms & User Terms:
- Aroma Descriptors
- Smell Notes
- Aroma Profile Terms
- What aromas were present
- Main aroma notes
Hierarchy:
- Parent Attribute ID: (blank) (Could be "Aroma Profile" if such a parent exists)
- Child Attribute IDs: (Individual aroma notes like "ATTR_AROMA_COFFEE_ROAST_001" could be children if managed this way, or they could be instances of this generic type mapped in Module C)
Measurement Details:
- Typical Scale Types:
  - [X] CATA (Check-All-That-Apply)
  - [ ] Frequency of Mention (derived from CATA)
- Typical Anchors: N/A for CATA (selection implies presence)
- Units: Presence (Yes/No), Percentage of Panelists Mentioning
Relationships:
- Related Statistical Concepts:
  - [X] Frequency
  - [X] Percentage
  - [X] Chi-Square (for comparing profiles)
- Influenced By Factors:
  - Product Ingredients
  - Processing (e.g., roasting, fermentation)
  - Storage Conditions
- Influences Attributes:
  - ATTR_AROMA_LIKING_001
  - ATTR_LIKING_OVERALL_001
Examples:
- List the top aroma notes for SLHB8.
- What were the main aroma notes for SLHB8 among millennials?
Keywords for Search:
- aroma
- smell
- descriptor
- profile
- CATA
- orthonasal
Notes for Pihu: When asked for "main aroma notes," typically refers to those with the highest frequency of mention from a CATA list. The specific list of notes is defined per study/product in Module C.
Created By: sensory_team_lead
Created At: 2023-01-18T11:00:00Z
Last Modified By: sensory_analyst_02
Last Modified At: 2023-10-25T16:15:00Z

2) Actual Data Example in JSON (for Module A - focusing on relevant attributes)¶

This JSON would contain entries for "Overall Product Liking," "Aroma Note (CATA)," and potentially other attributes that might come up in similar queries (like "Aroma Intensity," "Coffee-Roast Aroma Note" if it's a common specific note, etc.).

{
  "moduleName": "Sensory Attribute & Domain Concept Lexicon",
  "version": "1.1",
  "lastUpdated": "2023-11-15T10:00:00Z",
  "attributes": [
    {
      "attributeId": "ATTR_LIKING_OVERALL_001",
      "canonicalName": "Overall Product Liking",
      "category": "Hedonic",
      "status": "Active",
      "definition": "The overall degree to which a respondent likes or dislikes a product, considering all its sensory aspects after evaluation.",
      "scopeNotes": "Holistic measure of preference, distinct from liking of individual attributes.",
      "perceptionEvaluation": {
        "mechanism": "Subjective, integrated judgment based on the complete sensory experience.",
        "standardEvaluationTechnique": "Rating on a hedonic scale (e.g., 9-point).",
        "evaluationContext": "Typically at the end of product evaluation."
      },
      "synonymsUserTerms": [
        "Overall Liking", "Product Liking", "Liking Score", "Hedonic Score", "Preference Score"
      ],
      "hierarchy": { "parentAttributeId": null, "childAttributeIds": [] },
      "measurementDetails": {
        "typicalScaleTypes": ["Category Scale (e.g., 1-9 Hedonic)"],
        "typicalAnchors": [
          { "scalePoint": 1, "label": "Dislike Extremely" },
          { "scalePoint": 5, "label": "Neither Like nor Dislike" },
          { "scalePoint": 9, "label": "Like Extremely" }
        ],
        "units": "Score (e.g., out of 9)"
      },
      "relationships": {
        "relatedStatisticalConcepts": ["Mean", "Median", "Top-2-Box Percentage"],
        "influencedByFactors": ["All individual sensory attributes", "Product concept", "Respondent preferences"],
        "influencesAttributes": ["ATTR_PURCHASE_INTENT_001"]
      },
      "examples": ["What is the overall liking for Sample A?", "Compare overall product liking of SLHB5 and SLHB8."],
      "keywordsForSearch": ["liking", "preference", "hedonic", "acceptance", "overall score"],
      "notesForPihu": "Key performance indicator. Often compared against benchmarks or other products.",
      "createdBy": "sensory_team_lead",
      "createdAt": "2023-01-10T10:00:00Z",
      "lastModifiedBy": "sensory_analyst_01",
      "lastModifiedAt": "2023-11-01T14:30:00Z"
    },
    {
      "attributeId": "ATTR_AROMA_NOTE_CATA_001",
      "canonicalName": "Aroma Note (CATA)",
      "category": "Olfactory",
      "status": "Active",
      "definition": "A specific, individual aromatic characteristic perceived orthonasally, identified using Check-All-That-Apply.",
      "scopeNotes": "Refers to presence/absence of specific aroma descriptors. Intensity might be separate.",
      "perceptionEvaluation": {
        "mechanism": "Detection of specific VOCs associated with the descriptor.",
        "standardEvaluationTechnique": "Panelists select applicable aroma terms from a list.",
        "evaluationContext": "Orthonasal olfaction, before tasting."
      },
      "synonymsUserTerms": [
        "Aroma Descriptors", "Smell Notes", "Aroma Profile Terms", "What aromas were present", "Main aroma notes"
      ],
      "hierarchy": { "parentAttributeId": null, "childAttributeIds": [] },
      "measurementDetails": {
        "typicalScaleTypes": ["CATA (Check-All-That-Apply)", "Frequency of Mention"],
        "typicalAnchors": [], // N/A for CATA
        "units": "Presence (Yes/No), % Panelists Mentioning"
      },
      "relationships": {
        "relatedStatisticalConcepts": ["Frequency", "Percentage", "Chi-Square"],
        "influencedByFactors": ["Product Ingredients", "Processing", "Storage Conditions"],
        "influencesAttributes": ["ATTR_AROMA_LIKING_001", "ATTR_LIKING_OVERALL_001"]
      },
      "examples": ["List the top aroma notes for SLHB8.", "What were the main aroma notes for SLHB8 among millennials?"],
      "keywordsForSearch": ["aroma", "smell", "descriptor", "profile", "CATA", "orthonasal"],
      "notesForPihu": "For 'main aroma notes,' use highest frequency from CATA. Specific note list is in Module C per study.",
      "createdBy": "sensory_team_lead",
      "createdAt": "2023-01-18T11:00:00Z",
      "lastModifiedBy": "sensory_analyst_02",
      "lastModifiedAt": "2023-10-25T16:15:00Z"
    }
    // Add other relevant attributes like "Coffee-Roast Aroma Note", "Sweet Aroma Note", etc.
    // if they are commonly queried or part of standard TagTaste lexicons.
  ]
}

3) JSON Structure (for Module A - Simplified Comments)¶

This is the generic structure. The actual content would be populated as shown in the example above.

{
  "moduleName": "Sensory Attribute & Domain Concept Lexicon",
  "version": "1.0",
  "lastUpdated": "YYYY-MM-DDTHH:MM:SSZ",
  "attributes": [
    {
      "attributeId": "STRING_UNIQUE_ID",         // Unique ID for this attribute
      "canonicalName": "STRING",                 // The main, official name
      "category": "ENUM_STRING",                 // e.g., "Hedonic", "Olfactory", "Gustatory"
      "status": "ENUM_STRING",                   // e.g., "Active", "Deprecated"
      "definition": "TEXT_AREA_STRING",         // What this attribute means
      "scopeNotes": "TEXT_AREA_STRING",          // Clarifications on what it covers or doesn't
      "perceptionEvaluation": {
        "mechanism": "TEXT_AREA_STRING",         // How humans perceive it
        "standardEvaluationTechnique": "TEXT_AREA_STRING", // How it's usually measured
        "evaluationContext": "TEXT_AREA_STRING"  // When/how it's evaluated (e.g., "orthonasal")
      },
      "synonymsUserTerms": ["STRING_SYNONYM_1"], // Other names users might use
      "hierarchy": {
        "parentAttributeId": "NULLABLE_STRING_UNIQUE_ID", // Link to a broader attribute
        "childAttributeIds": ["STRING_UNIQUE_ID_CHILD_1"] // Link to more specific attributes
      },
      "measurementDetails": {
        "typicalScaleTypes": ["ENUM_STRING_SCALE_TYPE_1"], // e.g., "9-point Category Scale"
        "typicalAnchors": [
          { "scalePoint": "NUMBER_OR_STRING", "label": "STRING" } // e.g., 1: "Dislike Extremely"
        ],
        "units": "NULLABLE_STRING"                 // e.g., "Score out of 9"
      },
      "relationships": {
        "relatedStatisticalConcepts": ["STRING_STAT_CONCEPT_1"], // e.g., "Mean"
        "influencedByFactors": ["STRING_FACTOR_1"],          // e.g., "Product Temperature"
        "influencesAttributes": ["STRING_UNIQUE_ID_INFLUENCED_1"]
      },
      "examples": ["STRING_EXAMPLE_1"],           // How it might be used in a sentence
      "keywordsForSearch": ["STRING_KEYWORD_1"],   // For easier lookup
      "notesForPihu": "TEXT_AREA_STRING",          // Special instructions for the chatbot
      "createdBy": "STRING_USER_ID",
      "createdAt": "YYYY-MM-DDTHH:MM:SSZ",
      "lastModifiedBy": "STRING_USER_ID",
      "lastModifiedAt": "YYYY-MM-DDTHH:MM:SSZ"
    }
    // ... more attribute objects
  ]
}

Next Steps for Your Team (for Module A):

Sensory Team:
- Review the Attribute Entry Form structure. Does it capture everything needed to define a sensory attribute comprehensively from TagTaste's perspective?
- Start populating this form (or a spreadsheet version of it) for the most common and critical sensory attributes TagTaste deals with (e.g., Overall Liking, key aroma categories, basic tastes, key texture attributes).
- Pay close attention to canonicalName, definition, synonymsUserTerms, and typicalScaleTypes/typicalAnchors as these are very important for Pihu.
Technical Team:
- Review the JSON Structure. Is it clear and implementable for storing and retrieving this lexicon?
- Consider how Pihu's "Query Understanding Agent" would query this module (e.g., lookup by canonicalName, search by synonymsUserTerms or keywordsForSearch).