Views:

Intent matching algorithm is designed as a row-based search model. For example, “credit” and “card” terms are added in both single and double form to “Credit Card” menu. In this scenario, the menu contains “credit”, “card”, and “credit card” rows. Although all of the rows would match with 100% match score if the given utterance contains both words, “credit card” row should be selected. The higher number of matching terms are selected. However, 

Related Products: Chatbot 2 (NLP+Orhestrator), Conversational IVR 1 (NLP+SR), Virtual Assistant 2 (SR+TTS+NLP+Orchestrator)