Abstract
Considering the lack of systematic reviews on e-government research, this study includes research categories, spatial structure, research paradigms and noteworthy future topics in the domain of e-government. We collect 142 keywords from 2646 papers published in the Web of Science from 2000 to 2019 as the study object. Then, we identify four research categories: (1) technology and modelling in e-government, (2) drivers of e-government development, (3) public management and (4) governmental management. From public and government perspectives, we outline the spatial structure which includes theories and practices research and conclude four research paradigms: (1) theoretical modelling and application, (2) e-government development, (3) status of public management and (4) status of governmental management. Finally, we develop a 3D spatial map to analyse noteworthy topics and explore the well-studied themes: government-to-citizen, government-to-government and government-to-business, and the under-studied themes: government-to-civil society organisations and citizens-to-citizens, which helps scholars research e-government roundly.
1. Introduction
E-government, using information technology, provides the public, enterprises and governmental agencies with efficient government services [1]. Scholars have conducted an extensive literature review to examine the status of e-government research, which offers ongoing space for research focus adjustment [2–5].
Our review of existing literature mainly focuses on specific topics in the domain of e-government research, such as the sustainable development of e-government governance [6], the application of blockchain architectures [7], some fundamental issues regarding democracy [8], the dissemination of e-government [9], the dissemination of e-participation in public administrations [10], and the role of e-government in sustainable public procurement [11]. While Yildiz [12] thoroughly reviewed the limitations of e-government research in terms of research technology, model and definition, gaps remain. For example, how is the spatial structure of e-government research distributed? What are the basic research paradigms in the domain of e-government? What are the hotspots of e-government research in recent years and potential opportunities for e-government research in the future? As these problems have not been well solved, our research first incorporates bibliometric methods for resolving these gaps in e-government.
We used K-core and multidimensional scaling (MDS) to create a keyword network of e-government research using the Net Draw programme. The results show four categories of research topics: (1) technology and modelling in e-government, (2) drivers of e-government development, (3) public management and (4) governmental management. Accordingly, we created a knowledge map to delineate the spatial structure of e-government research and analyse its four-dimensional distribution, consisting of theoretical study, practical application, public perspectives and governmental perspectives. Then, to clarify each research topic, we discuss various research paradigms separately. The results show that the research topics primarily involve four research paradigms: (1) theoretical modelling and application, (2) e-government development, (3) the status of public management and (4) the status of governmental management. Finally, to find new hotspots for research, we draw up a 3D space map of e-government research based on issued ratio, theoretical studies and practical applications. The results suggest a lack of research on the themes of government-to-civil-society-organisations (G2CS) and citizens-to-citizens (C2C).
In addition to theoretical contributions, the novelty of our study is that we provide valuable insights and suggestions on knowledge about the potential relationships between current research topics and blind spots in that research and encourage more groundbreaking work in this field.
2. Data resources
For the journal in the Web of Science from 2000 to 2019, we chose ‘e-government’ as the topic, restricted type of document. We obtained 2793 documents, and there were 147 papers missing abstracts that could represent the author’s point of view, research results and specific observations. Therefore, we removed these and picked 2646 papers as our experimental samples. To conduct the analysis, we selected five items – author, title, keywords, abstract and year – to create data publishing forms, as shown in Table 1.
Structure of the extracted data records.
First, to be able to know what scholars are studying, we measured the contribution of the first author to the total number of citations in the Science Web. Second, we used Python 3.7 to measure the keyword frequency and selected keywords whose frequency was no less than 10, which included 507 keywords. From Table 2, we can see that the highest frequency is 2081 times, which belongs to the keyword ‘e-government’, the core keyword in our research sector. There is also a noticeable frequency for the keywords ‘information’, ‘government’, ‘management’, ‘public’ and ‘technology’, which are closely linked to ‘e-government’.
Frequency of keywords.
ICT: information and communications technology.
Finally, we counted the number of papers published each year, which is shown in Figure 1. The number of e-government papers rises gradually to a peak in 2016 and has since been declining. According to the Report of the United Nations E-Government Survey 2014: E-Government for the Future We Want, the government has started to concentrate on open data and address plans for open government data, opportunities and challenges. This represents a new trend in e-government research. Thus, many scholars have attempted to explore this new field, which led to the boom in e-government research after 2014. However, because of the lack of creative methods and technology, research on e-government is at a bottleneck stage. Hence, the number of papers declined slightly since 2016, and scholars have struggled with research to preserve the sustainable development of e-government.

Number of papers published in the proceedings of the Web of Science database in the period between 2000 and 2019.
3. Data analysis
Based on the above results, we selected co-word analysis as a method for evaluating theme relations to draw up our research framework. By analysing the trend of academic terms or words frequently appearing in the same literature, we analysed the structure of e-government research.
We extracted the co-occurrence of the 142 top keywords in 2646 papers in the Web of Science database and used CO_CO1.9 software to set up a two-keyword co-occurrence matrix M, where M∈R142 × 142, as shown in Table 3. Co-occurrence matrix M is a single-mode network matrix, in which each data point reflects how frequently each pair of keywords occurs simultaneously in the same document. The number 40 shown in the intersecting location between the second column and the second row in Table 3, for example, shows that the keywords of ‘e-government’ and ‘social media’ appear 40 times simultaneously in the same documents. We also removed the data elements on the diagonal, as they reflect the frequencies of their own co-occurrence.
Keywords co-occurrence matrix M (fragment).
Furthermore, we used social network analysis (SNA) to divide the 142 keywords into 10 subgroups based on the results, whose node degrees are 5, 7, 8, 9, 10, 11, 12, 13, 14 and 15. The node degree of 60% of the keywords is more than 10 times, reflecting the most work on topics by scholars. In addition, we used MDS to evaluate the research topics to determine the spatial positions of these 142 keywords, attempting to determine the study-oriented classifications and evolutionary paradigm. Then, we visualised the network of keywords and the effects of K-core and MDS analysis using Net Draw software, with the result shown in Figure 2.

Results of K-core and MDS on the keyword network of e-government research.
4. Discussion
4.1. Four categories of research topics
In Figure 2, we split the 142 top keywords into four major topics by keyword type. First, the 28 keywords at the bottom of this figure may be summed up as the topic of technology and model in the e-government. This topic could be further broken down into two subgroups: (1) software for the advancement of e-government: ‘information technology’ and ‘information and communication technology’; (2) model of growth for e-government: ‘technology acceptance model’ and ‘maturity model’. The main aim of this topic is to inform us about the latest common technologies and models in e-government research.
Second, in the middle of Figure 2, we summarised 54 keywords as the topic of drivers of e-government development. This topic is 2D: (1) system-driven e-government: ‘local government’ and ‘open government’; (2) technology-driven e-government: ‘digital government’ and ‘m-government’. This topic intends to reveal the factors that drive the development of e-government.
Third, the 41 keywords on the left of Figure 2 may be summed up as the topic of public management. This topic has two key aspects: (1) micro perspectives: ‘public service’ and ‘public value’; (2) macro perspectives: ‘public policy’ and ‘public administration’. This topic seeks to examine the effect of public administration.
Finally, we sum up the remaining 19 keywords on the right of Figure 2 as the topic of government management in each aspect. There are two subgroups to this topic: (1) governance types: ‘project management’, ‘knowledge management’ and ‘information management’; (2) government management features: ‘security’ and ‘privacy’. This topic intends to examine further the main aspects of government management and e-government management features.
4.2. Spatial structure of e-government research
In terms of the research categories in the domain of e-government, we find that the spatial structure of e-government research is widened by the use of MDS in networks of keywords. In Figure 3, ‘e-government’ is located in the centre because it is the central keyword in the field of e-government. We then divided the remaining 141 keywords into four quadrants.

Spatial structure of e-government research.
From theoretical study to practical application, we developed space maps organised both horizontally, which reflect the entire process flow of e-government research and longitudinally from government to the public. Because Yildiz [12] stated that the paper contained five e-government models: government-to-government (G2G), government-to-citizen (G2C), government-to-business (G2B), government-to- civil-society-organisations (G2CS) and citizen-to-citizen (C2C), and the results of this study showed that scholars focus on the G2C, the keywords in the first and second quadrants refer to study themes on public perspectives.
From a horizontal point of view, the first quadrant contains public-perspective activities, represented by keywords, such as ‘public management’, ‘information and communication technologies’, ‘public participation’, ‘service delivery’, etc. The keywords in the second quadrant, such as ‘public policy’, ‘regulation’ and ‘public services’, refer to public-perspective theories. The keywords in the third and fourth quadrants relate to the research themes on government perspectives. Keywords ‘open government’, ‘local government’, ‘e-government’, ‘m-government’, etc., reflect government-perspective ideas, and government-perspective activities include keywords, such as ‘e-services’, ‘e-participation’ and ‘e-readiness’.
From a vertical point of view, we constructed space maps that communicate from a government and public viewpoint. The second and third quadrants form the core of e-governmental theory. The first and fourth quadrants include the steps and means that constitute a connection between government and the public. Combining Figures 2 and 3, we suggest that the spatial framework of e-government research should include four parts: theories of public perspective, theories of government perspective, practices of public perspective and practices of government perspective.
4.3. Specific research paradigms
In this section, we conclude the different research paradigms according to the four research topic categories.
First, from Figure 2, we consider ‘the technology acceptance model’ [13–15] and ‘information and communication technologies’ [16–18] as the methods most studied in the field of e-government. Scholars use these methods specifically to tackle a particular e-government problem or to study how to refine these methods. Thus, according to the keyword distribution, the first research paradigm of e-government primarily focuses on theoretical models and applications.
Second, searching the literature using the keyword ‘local government’, we found the top five countries to be studied were Spain [19,20], Europe [21], the United Kingdom [22,23], Italy [24] and Indonesia [25]. Similarly, we found the open government first listed by McDermott in 2010 [26]. Through the study of keyword distribution, we found that the subsequent open government research focuses on two perspectives: open government itself [27,28] and partnerships between open government and open data [29–32]. Meanwhile, technology and big data are pushing governments to develop two new forms: digital government and mobile government (also called m-government). Research on digital government [33,34] and m-government [35–37] has been conducted. Nevertheless, there is comparatively less research on m-government than other topics, so we propose to extend m-government research. As a result, the second research paradigm for e-government focuses on e-government development.
Third, using Python 3.7 to quantify the frequency of keywords about the public, we found the keywords from high to low frequency, ‘public administration’, ‘public sector’, ‘public management’, ‘public services’, ‘public values’, ‘public policies’ and ‘public participation’, were the most common. Strejcek and Theil [38] first used ‘public administration’ as a term, which reflects the point of departure for public science. Papers that focused on contact between the public and the government consumed almost their entire study. In this case, researchers aimed to research public services to improve public participation [39,40]. Nevertheless, these studies ignore the public-to-public relationship, which is a novel aspect that needs to be discussed. We assume that the research paradigm for this topic focuses mainly on the status of public management according to the above debate.
The development of e-government naturally involves collaboration and efficient national, state and local government and government agency management. On the last point, scholars have studied e-government management, using the keywords ‘project management’ [41], ‘change management’ [42], ‘information management’ [43], ‘knowledge management’ [44] and ‘records management’ [45]. Nevertheless, as mentioned above, public administration is restricted to interactions between the public and government. Thus, as the public is a significant participant in the whole cycle of e-government, scholars should turn their attention to public-to-public internal management. From this, the study’s fourth research paradigm focuses primarily on the status of government management.
4.4. Noteworthy topics in the future
Combining Figures 2 and 3 shows that prior research paradigms affected almost all areas, including theories of e-government, practices of e-government, government studies and public application. Such research has studied the advancement of e-government in depth. The sustainable growth of e-government, the application of the new approaches to e-government study and local government services are brought into research. Filgueiras et al. [46] proposed digital transformation in governments as a phase of systemic change in public institutions, concerning the position of agents, the contexts of choices and the factors that explain the decision to digitise public services. The level of public engagement in government affairs clearly represents the success of e-government. Accordingly, research on public engagement will become one of the most valuable subjects of e-government research. Nonetheless, in Figures 2 and 3, we find only a few keywords that belonged to the study of public engagement, which means that researchers have not paid adequate attention to public engagement. In addition, the factors affecting public engagement are important for e-government research. There are still a limited number of scholars who are concerned about the role of public participation in e-government research, such as Mukamurenzi et al. [47] and Tassabehji et al. [48].
To dig deeper into new points in Figure 3, this study defines a new spatial structure with 3D coordinates: issued amount, theoretical study and practical application, as shown in Figure 4. Each plane represents a subject for study. The area of a circle that is projected in a straight line is the number of posts regarding this pattern. We divided the graph into two groups based on the amount of work done: (1) well-studied themes (A, B and C): G2C (47.66%), G2G (43.39%) and G2B (4.80%); (2) under-studied themes (D and E): G2CS (3.67%) and C2C (0.49%).

3D space map of e-government research.
Therefore, in the future, it will be valuable for scholars to focus on C2C, including the following themes: how public value manifests in the development of government affairs, how to construct a suitable open platform for local governments, how to enhance data transparency for the public and how to improve government service for the public. Then, if governments and the public cooperate well, scholars should consider relationships among the public, such as how they share information and how they interact on government-public platforms like government microblogs. It will also be worth focusing on G2CS, such as how civil society organisations influence e-government and how to improve the relationship between civil society organisations and governments.
5. Conclusion
From the Web of Science database from 2000 to 2019, we selected the 142 top keywords from 2646 articles on e-government science. Based on these keywords, we mapped a keyword network and its spatial structure from these top research topics and paradigms using SNA’s K-core, co-word analysis, MDS and the UCINET 6.0 and Net Draw software packages. The results of the study show that there are four key categories in e-government research: (1) technology and model in the e-government, (2) drivers of e-government development, (3) public management and (4) government management. In illustrating the spatial structure of e-government research using a knowledge map, the horizontal axis displays the research themes on theoretical studies or practical applications in e-government, and the vertical axis displays the research themes from the viewpoint of the public or government. Meanwhile, the findings indicate that the research topics mainly include four research paradigms: (1) the theoretical models and application, (2) the form of e-government development, (3) the status of public management and (4) the status of government management. Finally, we identified future research topics: (1) processes between the public and government, specifically factors, such as the feeling of the public about the government; (2) the model of e-government, such as G2CS and, particularly, C2C.
5.1. Theoretical implications
Several theoretical contributions have been made to this analysis. At present, several researchers have suggested frameworks for solving e-government problems, and they have gradually brought analytical expertise from various perspectives into the field of e-government research. There are, however, some insights into the spatial structure of e-government research to expose the basic distribution of the study paradigm. Our analysis resolves the problem well. It acts as a tool to help e-government researchers explain the overall study and knowledge map structure of e-government. Therefore, this study offers a general context for providing insights into the future of e-government analysis. In the meantime, we integrated the knowledge map into e-government to present the spatial structure of e-government research intuitively and visually.
5.2. Practical implications
We used co-word analysis and related methods of visualisation to conduct in-depth thematic analysis, systematically and intuitively uncover the subject structure and development status and make predictions about future subject development. From the findings in this study, scholars can understand the current status of research. The implementation of multidimensional scale analysis also encourages the findings to be more intuitive, which is useful for researchers to understand the structure of e-government research. However, the growing number of noteworthy topics is helping scholars research e-government in depth. Ultimately, this study has value as a reference for scholars in e-government studies.
5.3. Limitations
This study has several limitations. Additional methods, such as analytical testing and bibliometric methods of document co-citation analysis and author co-citation analysis, should be used to validate this exploratory work, which can also provide different dimensions for study. In the meantime, we will further examine and continue to study this material.
Footnotes
Author contributions
H.J. made contribution to the study conceptualisation, writing, reviewing and editing; S.W. prepared the original draft; J.Y. contributed to the methodology.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by the National Natural Science Foundation of China (Grant Nos. 71704153, 71701180, 71704020 and 71801187) and China Postdoctoral Science Foundation (Grant No. 2018M642472).
