Document Type



This paper proposes an intelligent interface integrated services environment for electronic commerce. It contains four main modules, i.e., dialog manager, EC transaction ontology, information collector, and data analyzer. The dialog manager manages the transaction history of the user to collect the user-specific information. It also provides different helping functions for the user. The EC transaction ontology stores the business type of electronic commerce in the system. It classifies the business process into eight categories, i.e., transaction type, ordering pattern, merchandise type, logistic type, customer category, transport vehicle, delivery type, and service type. It also provides the relations and constraints between different business steps. The information collector instantiates and dispatches different information agents to extract, filter, and synthesize relevant transaction data from different data sources timely. The data analyzer is the main component of the system. It instructs the dialog manager to provide proper helping functions for the different users. It then uses the user model to classify the knowledge and skill level of a user. It also predicts the goal of the user by using case-based reasoning and fuzzy Petri-Net to recommend the user with the best transaction plan and browsing behavior. Finally, it monitors the quality of service that the system provided for the user. This environment not only provides an integrated human-machine interface but also manages the business information thoroughly for the users.