When customers are interested in a service or intend to buy it, they sometimes have questions on that service. In this study, we considered an inquiry system in which customers ask questions on a specific service and obtain correct information on the service. For such an inquiry system, a question-answering (Q&A) technology is needed. Many programming modules for such a technology have been developed and can be easily used for system development. In many Q&A technologies, machine-learning techniques are involved, and we need to prepare training data consisting of pairs of an answer and assumed questions. For training-data preparation, an answer set for a service should be defined as the first step and the answer set should cover all the information on the service that customers may ask about. By using a customer-behavior model and introducing a service-function model, we propose a method of effectively collecting knowledge information for an answer set on a service. Through a case study, we show that we can collect exhaustive knowledge information for an answer set with our method compared to the case in which domain experts collect knowledge information in their own way. For an actual project, we also considered an actual inquiry-system-development project, with training data obtained with the proposed method, and showed that the system covers almost all the information on the service that customers may ask after a user test.

Introduction

With the recent rapid expansion of artificial-intelligence (AI) technologies, many machine-learning-based programming modules such as text classification and image recognition, have been developed and made available as application programming interfaces (APIs). Developers do not have to worry about the details of the machine-learning algorithms but use the module function by just preparing the training data required for the machine-learning programming module. When developing an application for a service provider, developers need to develop training data for each application with the help of domain experts of the service provider. The output of a machine-learning-based system is probabilistic, and the output may change dramatically when the input slightly changes. Therefore, not only the application interface but also the output quality of the machine-learning component is important. The quality of output mainly depends on the training data, and training data should be developed carefully. However, there are few methods on training-data development that can be applied in a wide area. In this study, we considered a situation we develop an inquiry system that involves customers asking questions about a specific service. The inquiry system automatically responds to an inquiry from a customer on behalf of the domain expert. Question and Answering (Q&A) is a technology of automatically answering a user’s question1 2 and is a core technology for an inquiry system. With a Q&A technology the query is first extracted from the question. After that, a set of answer candidates is found from the answer collection with this query. Finally, a relevance score is assigned to each candidate for the output. An inquiry system sends the input text to the engine implementing the Q&A technology and obtains a set of answer candidates. From the candidates, the system will show a pre-specified number of answers with high scores as output. There are implementations on the Q&A engine that are published as an API. In this paper we consider a Q&A approach based on a text-classification technology．