Deep Neural Network-based Question Answering System for Customer Relationship Management

Authors

Chien-Hung Chen, Graduate Institute of Networking and Multimedia, National Taiwan UniversityFollow
Yung-Chun Chang, Taipei Medical UniversityFollow
Chih Hao Ku, Cleveland State UniversityFollow

Paper Number

1417

Abstract

Customer Relationship Management (CRM) is a crucial component of business management, while a question answering system can help users seek answers and advice efficiently. In this study, we develop a deep learning-based question answering system that can be integrated into CRM. We collected 10,443 questions and answers on TripAdvisor from 44 U.K. hotels and used natural language understanding and generation techniques to develop an intent-based question response model. Our proposed method outperforms baseline comparisons by achieving an F1-score of 87.34% in intent detection tasks and above-average performance in response generation through a survey study with 63 responses.

Comments

Paper Number 1417

Recommended Citation

Chen, Chien-Hung; Chang, Yung-Chun; and Ku, Chih Hao, "Deep Neural Network-based Question Answering System for Customer Relationship Management" (2022). PACIS 2022 Proceedings. 62.
https://aisel.aisnet.org/pacis2022/62