Exploiting Meta-cognitive Features for a Machine-Learning-Based One-Shot Group-Decision Aggregation

Hilla Shinitzky (), Dan Avraham, Yizhak Vaisman, Yakir Tsizer, Yaniv Leedon and Yuval Shahar ()
Additional contact information
Hilla Shinitzky: Ben-Gurion University of the Negev
Dan Avraham: Ben-Gurion University of the Negev
Yizhak Vaisman: Ben-Gurion University of the Negev
Yakir Tsizer: Ben-Gurion University of the Negev
Yaniv Leedon: Ben-Gurion University of the Negev
Yuval Shahar: Ben-Gurion University of the Negev

Group Decision and Negotiation, 2024, vol. 33, issue 1, No 5, 87-111

Abstract: Abstract The outcome of collective decision-making often relies on the procedure through which the perspectives of its members are aggregated. Popular aggregation methods, such as the majority rule, often fail to produce the optimal result, especially in high-complexity tasks. Methods that rely on meta-cognitive information, such as confidence-based methods and the Surprisingly Popular answer, have succeeded in various tasks. However, there are still scenarios that result in choosing the incorrect answer. We aim to exploit meta-cognitive information and learn from it, to enhance the group’s ability to produce a correct answer. Specifically, we propose two different feature-representation approaches: Response-Centered feature Representation (RCR), which focuses on the characteristics of the individual response, and Answer-Centered feature Representation (ACR), which focuses on the characteristics of each of the potential answers. Using these two feature-representation approaches, we train machine-learning models to predict the correctness of a response and an answer. The trained models are used in our two proposed aggregation approaches: (1) The Response-Prediction (RP) approach aggregates the results of the group’s votes by exploiting the RCR feature-engineering approach; (2) The Answer-Prediction (AP) approach aggregates the results of the group’s votes by exploiting the ACR feature-engineering approach. To evaluate our methodology, we collected 2514 responses for different tasks. The results show a significant increase in the success rate compared to standard rule-based aggregation methods.

Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10726-023-09855-9 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:grdene:v:33:y:2024:i:1:d:10.1007_s10726-023-09855-9

Ordering information: This journal article can be ordered from
http://www.springer.com/journal/10726/PS2

DOI: 10.1007/s10726-023-09855-9

Access Statistics for this article

Group Decision and Negotiation is currently edited by Gregory E. Kersten

More articles in Group Decision and Negotiation from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().