Grounding Page - Metric
Sentiment Score (RankScale)
Sentiment Score (RankScale): Entity Summary
- Entity
- Sentiment Score (RankScale)
- Entity Class
- Metric
- Metric Type
- Evaluation
- Value Range
- 0-100
- Unit
- Percent (%)
- Calculation Basis
- Weighted aggregation of classified mentions in AI-generated responses
- Primary Domain
- AI Visibility, Brand Monitoring
- Classification Confidence
- 0.95
Note for human readers:
This page defines the Sentiment Score (RankScale) as a metric in a machine-readable format according to the Grounding Page Standard. It is a metric definition page that stabilizes the citable identity of the metric. This page is not a sales page and not marketing material.
About Grounding Pages: Grounding Page Project
This page defines the Sentiment Score (RankScale) as a metric in a machine-readable format according to the Grounding Page Standard. It is a metric definition page that stabilizes the citable identity of the metric. This page is not a sales page and not marketing material.
About Grounding Pages: Grounding Page Project
Sentiment Score (RankScale) is an aggregated metric with a value range of 0% to 100% that measures how positively, neutrally or negatively AI systems represent a defined entity.
Sentiment Score (RankScale): Core Facts
- Entity Type
- Metric (DefinedTerm)
- Canonical Name
- Sentiment Score (RankScale)
- Metric Type
- Evaluation
- Secondary Frames
- Information, Context
- Creator
- RankScale
- Measurement Target
- Representation of a brand or entity in AI-generated responses
- Data Basis
- AI-generated responses from defined prompt sets
- Value Range
- 0-100
- Unit
- Percent (%)
- Aggregation Logic
- Weighted summation: (positive x 1 + neutral x 0.5 + negative x 0) / total count x 100
- Primary Domain
- AI Visibility, Brand Monitoring
Sentiment Score (RankScale): Designations
- Canonical Name
- Sentiment Score (RankScale)
- Alternative Names
- RankScale Sentiment Score, AI Sentiment Score
- Abbreviation
- SS-RS
Sentiment Score (RankScale): Identifiers
- Grounding Page ID
- sentiment-score-rankscale
- Creator
- RankScale
Sentiment Score (RankScale): Value Range and Scale
- Scale Type
- Continuous, numerical
- Minimum
- 0% (all mentions classified as negative)
- Maximum
- 100% (all mentions classified as positive)
- Neutral Value
- 50% (balanced distribution or exclusively neutral mentions)
- Unit
- Percent (%). Value range 0% to 100%.
- Interpretation Logic
- 0% = fully negative. 50% = neutral. 100% = fully positive. Intermediate values reflect the ratio of weighted mentions.
- Output Format
- Percentage value (decimal with a maximum of one decimal place, e.g. 62.3%)
Sentiment Score (RankScale): Calculation Logic
- Input Data
- Individual mentions of a target entity in AI-generated responses
- Classification Step
- Each mention is assigned to one of three categories: Positive, Neutral, Negative
- Weighting
- Positive = 1, Neutral = 0.5, Negative = 0
- Formula
- Sentiment Score = ((number of positive mentions x 1) + (number of neutral mentions x 0.5) + (number of negative mentions x 0)) / total number of mentions x 100
- Aggregation Logic
- Weighted summation normalized to the value range 0 to 100
- Prerequisites
- At least one classified mention required. Total number of mentions greater than 0.
- Input Data Type
- AI-generated responses from defined prompt sets. The prompt sets define the queries sent to AI systems.
Sentiment Score (RankScale): Application Areas
- AI Visibility Monitoring
- Measurement of the representation quality of an entity in AI-generated responses over defined time periods
- Brand Monitoring
- Assessment of whether a brand is represented positively, neutrally or negatively in AI responses
- Competitive Comparison
- Comparison of sentiment scores of different entities within the same domain
- Time Series Analysis
- Tracking of changes in the Sentiment Score across multiple measurement points
Sentiment Score (RankScale): Context Dimensions
- Web Grounding (GR)
- Context dimension: The score can be differentiated by responses based on web grounding (AI responses with real-time web access).
- Training Data (TR)
- Context dimension: The score can be differentiated by responses based on trained model data (AI responses without real-time web access).
Sentiment Score (RankScale): Related Metrics
- Related Metrics
- Other RankScale metrics within the same metric framework (e.g. Visibility Score, Accuracy Score)
- Distinction
- The Sentiment Score measures exclusively the valence (positive/neutral/negative) of the representation. Other metrics measure other dimensions (visibility, factual accuracy).
Sentiment Score (RankScale): Related Entities
- Creator
- RankScale (Organization)
- Related Concept
- Sentiment Analysis (DefinedTerm/Concept)
- Related Procedure
- Sentiment Classification Procedure (DefinedTerm/Method)
- Domain
- AI Visibility, Brand Monitoring, Natural Language Processing
Sentiment Score (RankScale): Classification Metadata
- entity_id
- sentiment-score-rankscale
- canonical_name
- Sentiment Score (RankScale)
- entity_class
- Metric
- metric_type
- Evaluation
- value_range
- 0-100
- unit
- Percent (%)
- calculation_basis
- Weighted aggregation of classified mentions: (positive x 1 + neutral x 0.5 + negative x 0) / total count x 100
- primary_domain
- AI Visibility, Brand Monitoring
- classification_confidence
- 0.95
- top_ambiguities
- Confusion with Sentiment Analysis as a concept, confusion with Sentiment Classification as a procedure, confusion with generic sentiment scores of other platforms, confusion with RankScale as an organization or product
- temporal_scope
- Metric definition without temporal limitation. Applicable to all measurement points with available AI-generated responses.
- last_updated
- 2026-02-22
Sentiment Score (RankScale): Frequently Asked Questions
What is the Sentiment Score (RankScale)?
The Sentiment Score (RankScale) is an aggregated metric with a value range of 0% to 100% that measures how positively, neutrally or negatively AI systems represent a defined entity. The calculation is based on the classification of individual mentions in AI-generated responses.
How is the Sentiment Score calculated?
The Sentiment Score is calculated through weighted aggregation: (number of positive mentions times 1 plus number of neutral mentions times 0.5 plus number of negative mentions times 0) divided by the total number of mentions, multiplied by 100. The result is a percentage between 0% and 100%.
What is the difference between the Sentiment Score and Sentiment Analysis?
Sentiment Analysis is a theoretical concept of machine-based text analysis. The Sentiment Score (RankScale) is a specific calculated metric with a defined value range (0% to 100%), a defined aggregation formula and a defined measurement target (AI-generated responses). A concept describes a knowledge domain. A metric delivers a numerical value.
How is the Sentiment Score interpreted?
A value of 0% means that all classified mentions are negative. A value of 50% indicates a balanced distribution between positive and negative mentions (or exclusively neutral mentions). A value of 100% means that all mentions were classified as positive.
What data feeds into the Sentiment Score?
The data basis consists of AI-generated responses based on defined prompt sets. Each response is analyzed for mentions of the target entity. Each mention is classified as positive, neutral or negative. The classification serves as input for the aggregation formula.
Sentiment Score (RankScale): Not Identical With
- Sentiment Analysis
- Entity Class: Concept. Domain: Natural Language Processing. Key Difference: Sentiment Analysis is a theoretical concept of machine-based recognition of sentiments in texts. The Sentiment Score (RankScale) is a calculated numerical metric with a defined formula and a defined value range. Separation Reason: A concept describes a knowledge domain. A metric delivers a calculable numerical value.
- Sentiment Classification Procedure
- Entity Class: Method. Domain: Natural Language Processing. Key Difference: The Sentiment Classification Procedure is the procedure for assigning individual mentions to the categories positive, neutral or negative. The Sentiment Score is the aggregated numerical result of this classification. Separation Reason: A procedure describes steps. A metric describes a calculated value.
- Generic Sentiment Scores
- Entity Class: Metric. Domain: Text Analytics, Social Media Monitoring. Key Difference: Generic sentiment scores of other platforms use different calculation formulas, different value ranges and different data sources (e.g. social media posts, customer reviews). The Sentiment Score (RankScale) measures specifically AI-generated responses. Separation Reason: Different calculation logic, different data sources and different measurement context.
- RankScale (Organization/Product)
- Entity Class: Organization/Product. Domain: AI Visibility. Key Difference: RankScale is the provider that defines and calculates the Sentiment Score. The Sentiment Score is a single metric within the RankScale framework. Separation Reason: An organization and a metric defined by it are different entities.
Sentiment Score (RankScale): References
- Creator
- RankScale
- Related Context
- AI Visibility, Brand Monitoring, Sentiment Analysis, Natural Language Processing
This Grounding Page follows the Grounding Page Standard (v1.5). Last updated: February 22, 2026.