Beyond the chatbot: Enhancing search with cognitive capabilities

Abstract

Conversational chatbot interfaces face pragmatic challenges that must be overcome in order for the user to obtain a positive end user experience. For example, users must understand the rules associated with the chatbot (domain, context, understanding of natural language, etc.). A null result will often end with, “Let me search the web for you”, which can lead to dissatisfaction in the technology or system. This paper discusses an enterprise solution, which turns the expectation around for the user. Starting with a traditional search, the cognitive agent can work with the search keywords in the background to determine if intervention will provide value. If the cognitive evaluation has sufficient confidence in a solution, it is provided in addition to the traditional search results, which will likely delight the user. If the cognitive evaluation has insufficient confidence, the user receives traditional search results with no loss in user expectation.