Multi-agent systems of large language models as weight assigners: An approach to collaborative weighting in spatial multi-criteria decision-making
Notice bibliographique
Résumé
The integration of artificial intelligence (AI) technologies in decision-making processes is gaining momentum. Specifically, Large Language Models (LLMs) and multi-agent systems (MAS) hold considerable potential for transforming the landscape of multi-criteria decision-making (MCDM), particularly in addressing challenges posed by complex, multifaceted group decision-making environments. Conventionally, the collaborative expert weighting approach has been instrumental in spatial MCDM to ensure the accuracy and robustness of decisions. However, this approach is often subject to biases, significant time consumption, and logistical challenges in expert aggregation. This paper explores the feasibility of employing MAS and LLMs as substitutes for group expert-based weighting mechanisms in spatial MCDM by introducing the Weight Assignment by LLM-based MAS (WALMAS) method. In this method, LLMs such as OpenAI GPT-4o, Google Gemini, and Microsoft Copilot were regarded as primary agents, with multiple decision-making agents such as environment, urban planning, geography, and social specialists considered as a substitute for human experts depending on the nature of the problem. In the MAS space, following the parsing and extraction of initial weights from LLMs, a two-level algorithm was developed. The first level of this method involved the removal of outlier weights using the interquartile range (IQR) method. The second level of the method involved gradual negotiation and reaching consensus in an iterative process based on Kendall's W index. The proposed method, grounded in a GeoAI framework, was evaluated through its application to the landfill site selection problem. The findings and sensitivity analysis demonstrated that this method facilitates the efficient and reliable weighting of criteria, while ensuring the convergence of weights. Additionally, an analysis was conducted to identify the similarities and differences between LLMs in terms of weighting, as well as to determine the most effective expert agents in weighting. The analysis of human experts' satisfaction with the proposed method was also evaluated as very promising. This research demonstrates the effectiveness of AI-based tools in enhancing decision-making efficiency, consistency, and adaptability across spatial planning contexts.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction distillée sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Apprise à partir de 10 348 étiquettes directes de Codex et de 10 348 étiquettes directes de Gemma. Le mode candidate est l'union des têtes enseignantes seuillées; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont ni des étiquettes humaines ni des étiquettes directes de modèles de pointe.
Scores Codex et Gemma par catégorie
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,008 | 0,010 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,000 |
| Méta-épidémiologie (sens large) | 0,001 | 0,000 |
| Bibliométrie | 0,002 | 0,004 |
| Études des sciences et des technologies | 0,000 | 0,000 |
| Communication savante | 0,001 | 0,001 |
| Science ouverte | 0,002 | 0,001 |
| Intégrité de la recherche | 0,000 | 0,000 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,000 | 0,000 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule tête enseignante, pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».