Natural Language Processing in IBM Watson Assistant, an Automatic Verification Process

摘要

The exponential growth of Artificial Intelligence powered systems affect us all. IBM Watson Assistant is one of the central AI-powered systems used in small and larger business. For instance, in Portugal, IBM Watson is powering call-center systems for companies in banking and telecommunications business. This paper proposes a strategy that can automate the verification process of generalization capability in the creation of chat-bots using IBM's platform. K-Fold cross-validation is a favorite technique in machine learning for estimating the performance of a learned hypothesis on a data set. Therefore, the proposed method is not new for testing. However, this method is newly applied to the chat-bot application using IBM's platform. In this paper, the primary goal is to make the chat-bot testing process automated, with the objective of making it faster, more productive, and efficient. Algorithms like k-fold cross-validation demonstrate the need for a representative and reasonable amount of data when it comes to training IBM Watson in his ability to learn.