A Social Sensing Approach for Quality Changes of Real-World Services

Qianqian Shu,Zhiying Tu,Xiaofei Xu,Zhongjie Wang
DOI: https://doi.org/10.1109/sose49046.2020.00017
2020-01-01
Abstract:In today's dynamic business environment, services keep changing and being updated to adapt to user demands evolution and technology trends continuously. Customers are eager for information about changes in the quality of services. So, they can make accurate decisions on service selection and get a better user experience. There are quite a lot of ways for service providers to actively promote their services, such as email notifications, mobile Apps notifications, real-world advertisements, etc. However, it is still difficult for customers to catch these changing information comprehensively and timely. Customers must have a quality changing sensing tool. Previous studies on quality sensing in services computing community are usually focused on simple web-based API services, and quality indicators to be sensed are usually technical-oriented. In this paper, we present a new approach for sensing service quality changes by social sensing. It makes use of publicly-disseminated news reports and user reviews on real-world services to identify what quality indicators are changing and which direction a change is towards (positive or negative). Deep learning is used for mining quality change information from a text corpus, and experiments conducted on real data of nursing home services show the effectiveness of the proposed method.
What problem does this paper attempt to address?