Sensory Evaluation

1
Sensory Evaluation

• Decision Making processes
• based on Sensory Evaluation

Luis Martínez López Computer Sciences Dpt..
University of Jaén Email martin_at_ujaen.es Website
http//wwwdi.ujaen.es/martin
2
INDEX
3
INDEX

• SENSORY EVALUATION
• DECISION MAKING
• BASED ON SENSORY EVALUATION
• APPLICATION
• TEXTILE PRODUCTS
• CONCLUSIONS

4
SENSORY EVALUATION
5
SENSORY EVALUATION

• Evaluation of Signals
• Sight
• Smell
• Taste
• Touch
• Hearing
• Mind filters and makes concious the signals and
  forms perceptions

6
SENSORY EVALUATION

• In some aspects human senses are still superior
  to many artificial devices.
• Business, Industry
• Human operators vs artificial devices
• Quality control
• Hazard detection
• Collision Prevention
• Etc.

7
SENSORY EVALUATION

• Human Perceptions
• Operational tasks (not only)
• Consensus in groups
• Designing products
• Deriving stimates
• Decision models
• Etc.

8
SENSORY EVALUATION

• Dealing with human perceptions
• In the past
• Statistical Methods
• Dont work too well
• Implies independence minimal sample size
• Not Satisfied in real situation
• Perceptional uncertainty cannot be modelled this
  way

9
SENSORY EVALUATION

• Dealing with human perceptions
• Recently
• Intelligent Methods
• Fuzzy Technology
• Neural Nets
• Evolutionary Computing
• They allow to model mathematically perceptions

10
SENSORY EVALUATION

• APPLICATIONS
• Quality Inspection of products
• Risk evaluation
• Safety Evaluation
• Investment evaluation
• Marketing Study

11
SENSORY EVALUATION

• COMMON FRAMEWORK
• Sensory panel
• Group of individuals
• Evaluating a number of samples
• Each panellist
• Provides a score for each evaluated descriptor
• Information modelling
• Numerical
• Linguistic

12
SENSORY EVALUATION

• USEFULNESS
• Quality Inspection
• Characteristics
• Foods
• Textile products
• Chemical products
• Cosmetic products
• Trained expert panel sistematically determined
• Results
• New products can be designed integrating the data

13
SENSORY EVALUATION

• LOOKING FOR
• MATHEMATICAL FORMULATION
• EFFICIENT WORKING PROCEDURE
• PERTINENT COMPUTING METHOD
• EVALUATION SCORES
• DIFFERENT DOMAINS (DIFFERENT BACKGROUND)
• CLASSICAL METHODS
• LOST INFORMATION

14
SENSORY EVALUATION

• BIBLIOGRAPHY
• Intelligent Sensory EvaluationMethodologies and
  ApplicationsRuan, Da Zeng, Xianyi (Eds.) 2004,
  XIII, 443 p. 129 illus., Hardcover
• Zeng X, Koehl L, Sanoun M, et al.Integration of
  human knowledge and measured data for
  optimization of fabric hand INTERNATIONAL
  JOURNAL OF GENERAL SYSTEMS 33 (2-3) 243-258
  APR-JUN 2004

15
DECISION MAKING
16
DECISION MAKING BASED ON SENSORY EVALUATION

• EVALUATION PROCESS
• DECISION MAKING
• ELEMENTS
• PANEL
• MULTIPLE EXPERTS
• MULTIPLE CRITERIA
• SCORES
• UNCERTAINTY (PERCEPTIONS)
• PREFERENCE MODELLING DIFFERENT

17
DECISION MAKING BASED ON SENSORY EVALUATION

• FRAMEWORK
• MEMC-DM
• CONTEXT (UNCERTAINTY)
• NUMERICAL
• LINGUISTIC
• DECISION PROCESS
• AGGREGATION
• EXPLOITATION EVALUATION gt RANKING

18
DECISION MAKING BASED ON SENSORY EVALUATION

• DECISION PROCESS
• AGGREGATION
• EXPLOITATION
• OVERALL VALUE
• RANKING

19
SENSORY EVALUATION

• BIBLIOGRAPHY
• F. Herrera, L. Martínez. P.J. Sánchez. Managing
  non-homogeneous information in group decision
  making. European Journal of Operational Research
  1661(2005) pp. 115-132
• L. Martínez, J. Liu, J.B. Yang,F. Herrera, A
  Multi-Granular Hierarchical Linguistic Model for
  Design Evaluation Based on Safety and Cost
  Analysis. International Journal of Intelligent
  Systems. 2005, Vol 20, Issue 12, pp.1161-1194.

20
APPLICATION
21
APPLICATION EVALUATING TEXTILE PRODUCTS

• PROJECT
• FUZZY QUALITY EVALUATION APPLIED TO TEXTILE
  PRODUCTS
• UE EQUAL
• WEBSITE
• http//wwwdi.ujaen.es/martin/scv/equal/principal.
  html

22
APPLICATION EVALUATING TEXTILE PRODUCTS

• PANEL
• MULTIPLE EXPERTS
• MULTIPLE CRITERIA
• NUMERICAL
• CRISP
• INTERVALS
• LINGUISTIC

23
APPLICATION EVALUATING TEXTILE PRODUCTS

• FRAMEWORK
• r paneles
• h(i) individuos
• conjuntos de términos
  lingüísticos
• conjunto de n muestras a
  ser evaluadas
• vector de utilidad para
  cada muestra

24
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM

25
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM
• Prototype v 1.0

26
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM
• Prototype v 1.0 Crear Encuesta

27
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM
• Prototype v 1.0 Definir dominio de información

28
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM
• Prototype v 1.0 Rellenar Encuestas

29
APPLICATION EVALUATING TEXTILE PRODUCTS

• EVALUATION SYSTEM
• Prototype v 1.0 Resultados

30
CONCLUSIONS
31
CONCLUSIONS

• SENSORY EVALUATION
• DIFFICULTIES
• UNCERTAINTY
• DIFFERENT TYPE OF INFORMATION
• NECESSITIES
• MATHEMATICAL FORMULATION
• EFFICIENT WORKING PROCEDURE
• PERTINENT COMPUTING METHODS
• IMPROVEMENTS
• FUZZY TOOLS
• APPLICATIONS
• QUESTIONNARIE