Depending on the amount of data to process, file generation may take longer.

If it takes too long to generate, you can limit the data by, for example, reducing the range of years.

Article

Download BibTeX

Title

From investigation of expressiveness and robustness to a comprehensive value-based framework for multiple criteria sorting problems

Authors

[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ 2 ] Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] employee | [ SzD ] doctoral school student

Scientific discipline (Law 2.0)

[2.3] Information and communication technology

Year of publication

2025

Published in

Omega

Journal year: 2025 | Journal volume: vol. 131

Article type

scientific article

Publication language

english

Keywords
EN
  • Multiple criteria decision aiding
  • Sorting
  • Model expressiveness
  • Recommendation robustness
  • Interactions
  • Non-monotonicity
Abstract

EN We adopt an experiment-oriented perspective to investigate two essential characteristics – expressiveness and robustness – of multiple criteria sorting methods. We focus on the approaches from the family of UTADIS, learning the parameters of a value-driven threshold-based model from the Decision Maker’s assignment examples. Even if the considered properties are crucial for the methods’ reliability and usefulness in real-world scenarios, their verification through explicit numerical tests has been so far neglected. On the one hand, expressiveness captures the models’ flexibility to reproduce different preferences, including simple and complex ones, meaningfully and accurately. On the other hand, robustness reflects the ability to deliver valid recommendations and ensure proper conclusiveness given the multiplicity of compatible preference model instances. We consider different variants of UTADIS, from assuming monotonic and preferentially independent criteria to more advanced settings that relax the monotonicity constraints or represent interactions. The experimental results capture the trade-off between the considered quality dimensions, indicating that richer models are characterized by greater expressiveness and lesser robustness. We also formulate a comprehensive framework indicating when some variant should be used, given the nature of supplied preferences or problem characteristics. These findings aid decision analysts in making robust recommendations in different contexts and help refine preference modeling assumptions. The framework’s practical use is illustrated in a case study involving sorting mobile phone models into pre-defined preference-ordered classes.

Date of online publication

15.10.2024

Pages (from - to)

103203-1 - 103203-25

DOI

10.1016/j.omega.2024.103203

URL

https://www.sciencedirect.com/science/article/pii/S0305048324001671

Comments

Article Number: 103203

License type

CC BY (attribution alone)

Open Access Mode

czasopismo hybrydowe

Open Access Text Version

final published version

Date of Open Access to the publication

at the time of publication

Ministry points / journal

140

Impact Factor

6,7 [List 2023]

This website uses cookies to remember the authenticated session of the user. For more information, read about Cookies and Privacy Policy.