An Intelligent Interactive Voice Response System for Banking Domain

Interactive Voice Response (IVR) Systems have been developed by enterprises to provide better services to the customers by allowing an automated machine to interact with humans through the use of voice or Dual Tone Multi Frequency (DTMF) keypad tones. But the IVR systems being developed today do not provide ease of use to the customers as they are monotonous. The present IVR systems lack dynamism, have very less degree of failure recovery and use APIs that are system dependent. In this paper we are proposing an Intelligent Interactive Voice Response System (IIVR) that generates a context-based dialogue using semantic web service composition and Expert System. This approach provides quality service to the customer by reducing the conversation time and by providing maximum information needed by the user. We use Java Expert System Shell (JESS) Rule Engine for generating the dialogue based on the context and thus optimizing the request-reply dialogue process between the service provider and consumer of the IVR systems. Furthermore, we use Ontology-based knowledge representation and we extract the necessary information from the ontology using semantic web services to further refine the dialogue generated by the IVR system.