Dark Mode
Chemical Indicators of Wine Quality
Food & Beverage Consumption
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
The data provides a critical foundation for applying regression techniques—such as Support Vector Machine, Multiple Regression, and Neural Networks—to model wine quality and taste perception. It serves as an essential tool for quality assessment and certification, helping the wine industry identify the most influential chemical factors in wine making. Furthermore, it supports the development of models useful for market stratification, setting prices, and executing targeted marketing strategies by modelling niche consumer tastes.
Columns
The product contains 12 attributes essential for chemical analysis and quality prediction:
- fixed acidity: Measures the major fixed acids present in the wine (mean 6.85, range 3.8 to 14.2).
- volatile acidity: Relates to the presence of acetic acid, which can lead to a vinegar taste (mean 11.2, range 0.08 to 965).
- citric acid: An acid used to add 'freshness' and flavour (mean 0.33, range 0 to 1.66).
- residual sugar: The amount of sugar left over after fermentation ceases (mean 6.39, range 0.6 to 65.8).
- chlorides: Reflects the salt content (mean 40.6, range 0.02 to 346).
- free sulfur dioxide: The free form of SO2 available in the wine (mean 35.3, range 2 to 289).
- total sulfur dioxide: The sum of free and bound forms of SO2 (mean 138, range 9 to 440).
- density: The weight of the wine relative to water (mean 114, range 0.99 to 999).
- pH: Indicates the level of acidity/basicity (mean 3.19, range 2.72 to 3.82).
- sulphates: A wine additive that can contribute to overall flavour (mean 0.49, range 0.22 to 1.08).
- quality: A sensory evaluation score (mean 5.88, range 3 to 9).
- alcohol: The percentage of alcohol content.
Distribution
The equivalent underlying data file,
winequality-white new.csv, is 300.89 kB in size. Across the analytical columns, the data contains 4,898 valid records. It is highly structured, featuring 12 columns, with 0% mismatched or missing values reported. Visualizations derived from this data can be generated in platforms such as Loocker Studio, PowerBI, and Tableau. The expected update frequency for the product is annually.Usage
This data is ideally suited for:
- Developing predictive models for human wine taste perception.
- Supporting and streamlining wine certification and quality assessment processes.
- Identifying the most influential chemical factors for improving wine making techniques.
- Applying advanced regression models, with demonstrated success using the Support Vector Machine method.
- Creating wine stratification models, helpful for establishing premium brands and setting appropriate pricing.
Coverage
The data focuses exclusively on vinho verde wine samples sourced from the northwest region of Portugal. It includes examples of both white and red wine.
License
CC0: Public Domain
Who Can Use It
- Oenologists and Quality Control Experts: To validate human tasting results against analytical data.
- Wine Producers: To identify process improvements that directly affect perceived quality.
- Data Scientists and AI Developers: To benchmark and train machine learning models for sensory prediction.
- Marketing Professionals: To model consumer tastes and refine target marketing campaigns.
Dataset Name Suggestions
- Vinho Verde Quality & Taste Predictor
- Portuguese Wine Certification Analytics
- Chemical Indicators of Wine Quality
Attributes
Original Data Source: Chemical Indicators of Wine Quality
QUALITYLoading...
Free
Download Dataset in CSV Format
Recommended Datasets
Loading recommendations...