Two-way negotiation for intelligent hotel reservation based on multiagent: The model and system

摘要

Multiagent-based systems have been widely used in hotel reservations. Previous studies, however, mainly use one-way negotiation for decisions, which means that agents represent travelers selecting from a list of hotels to decide which room to reserve, while hotel providers cannot communicate sales strategies with travelers according to real-time situations to maximize profits. To address such problems, this paper proposes a Multiagent-Based Two-Way Negotiation for Hotel Reservation (MAB-TNHR) with three kinds of agents: agents representing the behavior of people (Tenant Agent and Landlord Agent), agents used for collecting data (Data Integrating Agent and Results Agent) and agents for system purposes (Controller Agent and Selector Agent) that together continuously observe the order requests and forecast their potential tenders to a number of qualified landlord agents based on the ontologies of collaborating participants. The traditional tenant-dominant reservation is converted into a form of tendering to implement this two-way negotiation, which means that Landlord Agents can actively respond to the selection by sending their tenders to suitable tenants and bargain on the price with the Tenant Agent to pursue their interests. In this paper, rules used in the application and a case study are presented to illustrate the implementation of MAB-TNHR. Verification shows that the MAB-TNHR can meet intricate and dynamic reservation demands automatically without the participations of tenants and landlords.