Multi-Criteria Multi-Participant Automated Negotiation: Belief Propagation-based Proposal Preparation and Real Time Opponent Learning
Notice bibliographique
Résumé
Automated negotiation has received considerable attention in the past few decades as a computer tool for modeling human negotiations. The aim of automated negotiation is capturing the model of interactions during the negotiation process and improving the efficiency and quality of real-world negotiations. Due to the complexity of negotiations, there are many challenges in modelling different aspects of this process. One of the important types of negotiation, which is the focus of this thesis, is multi-issue multi-participant argumentation-based negotiation. In such negotiations, several participants with different viewpoints and perspectives negotiate over several criteria. The involved parties in these negotiations exchange proposals (a set of values assigned to negotiation issues) and receive their opponents’ evaluation of the offered proposals as well as possible arguments. The primary goal of the negotiation is finding a solution that can satisfy all the involved parties. However, in multi-issue multi-participant negotiations, finding such a solution can be quite challenging because: 1- participants have different and, sometimes, conflicting preferences about the negotiation issues; and 2- these preferences are not usually revealed to others. The higher the number of negotiation issues (i.e., the dimensions of the search space for a satisfactory solution), the higher the number of unknown preferences and therefore, the harder to reach an agreement. Therefore, the negotiation process can take a long time before approaching a possible agreement. The current thesis studies two critical aspects of automated negotiation: proposal preparation and opponent modelling. The order of the offered proposals in consecutive rounds of the negotiation directly impacts the pace of reaching an agreement. Therefore, selecting the right proposal for each round based on the interactions in the previous rounds is the key to effective negotiation. In this thesis, a novel proposal-preparation solution is proposed. It represents the negotiation issues and participants’ preferences via a graphical model and applies belief propagation to optimize this graph, the output of which is a proposal to offer to the participants. The thesis also discusses the problem of unknown preferences of the participants in this negotiation context. A recursive Bayesian filtering algorithm is proposed to learn/estimate the preferences of the opponents only through the limited information they exchange as the negotiation proceeds. The proposed approaches are then applied to two case studies to investigate their impact on the efficiency of the negotiation process. The experimental results show that using the presented proposal preparation and opponent modelling techniques, the efficiency of the negotiation process is increased by up to 85% in both case studies.
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 machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,010 | 0,028 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,001 |
| Méta-épidémiologie (sens large) | 0,002 | 0,001 |
| Bibliométrie | 0,001 | 0,001 |
| Études des sciences et des technologies | 0,001 | 0,002 |
| Communication savante | 0,003 | 0,005 |
| Science ouverte | 0,006 | 0,003 |
| Intégrité de la recherche | 0,003 | 0,003 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,004 | 0,001 |
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 source (Gemma direct ou Codex distillé), 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 ».