Data-Driven Understanding of Smart Service Systems Through Text Mining

Chiehyeon Lim, Paul P. Maglio
2018 Service Science  
Smart service systems are everywhere, in homes and in the transportation, energy, and healthcare sectors. However, such systems have yet to be fully understood in the literature. Given the widespread applications of and research on smart service systems, we used text mining to develop a unified understanding of such systems in a data-driven way. Specifically, we used a combination of metrics and machine learning algorithms to preprocess and analyze text data related to smart service systems,
more » ... luding text from the scientific literature and news articles. By analyzing 5,378 scientific articles and 1,234 news articles, we identify important keywords, 16 research topics, 4 technology factors, and 13 application areas. We define "smart service system" based on the analytics results. Furthermore, we discuss the theoretical and methodological implications of our work, such as the 5Cs (connection, collection, computation, and communications for co-creation) of smart service systems and the text mining approach to understand service research topics. We believe this work, which aims to establish common ground for understanding these systems across multiple disciplinary perspectives, will encourage further research and development of modern service systems. 155 (San Román et al. 2011). A search for "{TOPIC: (smart service system)}" in the Web of Science generates more than 5,000 results across engineering, computer science, information systems, control, transportation, healthcare, and other fields. Yet despite widespread application and the importance of research in this field, to our knowledge, in-depth understanding of such systems is still lacking in the literature. A unified understanding of smart service systems across different fields may facilitate development and innovation. Furthermore, such understanding would promote the use, integration, and improvement of technologies from a broad and application-oriented perspective. Specifically, a generic definition or representation of a smart service system will promote mutual understanding among people of different backgrounds, thereby facilitating collaborative analysis and the development of such systems. Similarly, a comprehensive categorization of applications related to smart service systems can contribute to our understanding of system variety and lead to the creation of synergy between different applications. However, such integrative work is not easy to achieve because of the variety and volume of studies and applications related to smart service systems. In this work, we develop an understanding of smart service systems by mining text related to these systems. The interdisciplinary body of text we analyzed includes scientific literature and news articles. The former discusses research topics and the technology factors of smart service systems; the latter describes application areas and business aspects. To capture the essence of the data, our analytics method uniquely incorporates metrics to measure the importance of the word-features of the data and unsupervised machine learning algorithms, such as spectral clustering (Von Luxburg 2007) and topic modeling (Blei et al. 2003 ). Our analysis of 5,378 scientific articles and 1,234 news articles identified significant keywords, research topics, technology factors (sensing, connected network, context-aware computation, and wireless communications), application areas, and a definition. Furthermore, we developed a conceptual framework of smart service systems and a hierarchical structure of smart service system applications by integrating our findings and those of existing studies. Establishing common ground for central concepts is essential for science (Boehm and Thomas 2013). To integrate perspectives and capabilities for developing smart service systems, we provide a systematized view of dispersed knowledge about smart service systems, integrating such knowledge into a robust conceptualization of these systems and identifying key research topics, technology factors, and areas of application. Our findings contribute to the theory and practice of smart service systems. Our work is unique in that it uses a data-driven approach to understanding these systems. In terms of methodological contribution, it is, to our knowledge, the first study to accomplish data-driven understanding of a service research topic. Our research methodology can be applied to other topics in the future. This paper is organized as follows. In Section 2, we review studies related to smart service systems and provide the conceptual foundations for this work. In Section 3, we describe the research methodology, including data collection and analysis methods. In Section 4, we describe the findings. In Section 5, we discuss the theoretical, managerial, and methodological implications of our work. In Section 6, we conclude with a discussion of future research issues.
doi:10.1287/serv.2018.0208 fatcat:6jkciu5kabgofatgxrr2b4tbxi