Big Data for Public Domain: A bibliometric and visualized study of the scientific discourse during 2000–2020

This article aims to investigate the trend of scientific publications under ‘big data and policy’ research during the last two decades, including the dynamics of the network structure of researchers and the institutions. Bibliometrics is utilized as a tool to reveal the dynamics of scientific discussions that occur through articles, published in international journals indexed/contained in the Scopus database; meanwhile, the analysis visualization is performed by using VOSviewer 1.6.16. The search results indicate that the United States serves as the country of origin for most productive author affiliations in publishing articles, the University of Oxford (United Kingdom) serves as the home institution for most productive author affiliations, and Williamson, B., from the University of Edinburgh (United Kingdom), is considered as the most prolific writer. In addition, the Swiss Sustainability Journal from MDPI is cited as the source for the most widely discussed publication topic in its journals. Further, ‘Big Data for Development: A Review of Promises and Challenges’ is regarded as the article with the most references. Additionally, the most discussed topics on ‘big data and policy’ include smart cities, open data, privacy, artificial intelligence, machine learning, and data science.


Introduction
The current public sector has been inseparable from the need for big data (Munne, 2016), and even big data has become a growing consensus among stakeholders, which emphasizes the importance of employing big data in governance worldwide and data sharing, Citizen sentiment analysis, Smart city and Internet of things (IoT) applications, Cybersecurity, and several algorithms have embodied tax data and people's online shopping trends, further analyzed and employed into innovations in government services to raise the effectiveness, efficiency, and satisfaction of the community. The big data revolution can present 'a better society' (Azzone, 2018) if big data can improve the quality of decision-making and decision-makers are aware of the potential and opportunities of using big data itself. Public policy today can be understood as a set of actions that affect the solution of a policy problem, namely dissatisfaction with a particular need, request, or opportunity for public intervention (Milano, 2014). The quality can be measured from the capacity to create shared value (Osborne, 2017).
The three dimensions are expected to improve public value (Azzone, 2018), accommodated by the utilization of big data, aiming: 1) to provide a personalized service, 2) to involve the final users in designing and producing a personalized service, and 3) to change the characteristics of the service over time. Further, big data enables decision-making in real time (Höchtl et al., 2016), which has the potential to revise the traditional model in each stage of the policy cycle and to replace it with a continuous evaluation model. Therefore, big data is pivotal, and it must be fulfilled by the government in accommodating each line of governance, which is in accordance with the public domain (Phillips & Higgott, 1999).
In its development, the discussion about the potential, benefits, and numbers of realities from practice in a number of countries provides an immense discourse in the presence of big data, which was previously regarded as a supporter of government administration; however, it gradually plays an important role in prompt decision making (P.B. . The study of big data and the public sector plays a pivotal role in comprehensively understanding how the scientific discussions are conducted in articles, published in leading international journals (on the Scopus database, the Web of Science (WOS), and a number of other indexers).
However, research that particularly examines trends in big data and policy publications has been inadequate. Recently, research merely reviews the evolution of "data science and big data" research (Raban & Gordon, 2020), Big data in marketing (Amado et al., 2018), interdisciplinary big data research (Hu & Zhang, 2017), and the evolution of big data research (Halevi & Moed, 2012). Therefore, the bibliometric analysis of big data and policies in scientific publications over the last twenty years (2000-2020) is deemed imperative.
The formulations of the problem in this research examine the trend of scientific publications in big data and policy research during the last two decades  and the dynamics of the network structure of researchers, institutions, and research areas in the field of big data and policy.
This study aims to provide quantitative and statistical analysis of publication trends over a wider range of timescales and to explore network structures and trends that have been established in the last 20 years. The utilization of bibliometric analysis is expected to address the gaps in research, involving the main actors such as the authors and institutions, and the countries and areas of research. In addition, this research is expected to provide global information and discourse from scientific publications that have been published in the last twenty years related to big data and policy.

Methods
Data source -This research is regarded as a bibliometric study, engaging the articles related to big data and policy, which were in the Scopus database from 2000 to 2020. The utilization of Scopus as a data reference is based on the consideration that Scopus indexed publications serve as the standard in scientific publications as widely used in bibliometric analysis (Falagas et al., 2008) PubMed remains an optimal tool in biomedical electronic research. Scopus covers a wider journal range, of help both in keyword searching and citation analysis, but it is currently limited to recent articles (published after 1995. The research framework is presented in Fig. 1.

Search strategy-
The search strategy is utilized to identify publications in the field of public policy by title or abstract. This study specifically utilizes the Scopus database as of December 10, 2020, navigated through: the search options (TITLE-ABS-KEY ("big data") AND TITLE-ABS-KEY ("policy") AND PUBYEAR <2021).
In addition, to provide a more comprehensive reference, this research limits the subject area into "Social Sciences", with the source type of: "journal", "conference proceeding", "book", and "book series", written only in English (obtaining the 938 articles).

Bibliometric and Visualized analysis
-Bibliometrics is utilized to navigate the development of science and technology through the production of entire scientific literature at a certain level of specialization (Okubo, 1997;Prakoso Bhairawa Putera et al., 2020).

Results and Discussion
The development of 'big data and policy'  Bibliometric Analysis Framework for 'big data and policy'

Source: Processed by Author
Bibliometric and Visualized analysis -Bibliometrics is utilized to navigate the development of science and technology through the production of entire scientific literature at a certain level of specialization (Okubo, 1997;Prakoso Bhairawa Putera et al., 2020). Such analysis is valuable to substitute: one country towards another in the global scope; an institution with other institutions within the scope of a country/between countries; and individual researchers/authors in relation to their communities (Prakoso Bhairawa Putera & Rostiena, 2021). Bibliometric data visualization is performed by VOSviewer (analysis software to map analysis based on keywords, authors, countries, and journals) (   Publication Trends 'big data and policy'

Contribution of Countries / Institutions / Authors
Based on the bibliometric metadata, it is revealed that there are 78 countries from which the authors' affiliations present the research theme of 'big data and policy', and the Top 10 countries are illustrated in     Table 1). Based on Table 1 Table 2.   When viewed from the Quality of paper, measured from the total citation / total paper produced, forms six clusters (see Figure 7). These results are in accordance with Table 3. Based on Table   3, it is apparent that the 25 sources of 'big data  (as many as 10.53% or only two journals), and Q3 (as many as 5.36% or one journal only). This indicates that the article 'big data and policy' is a topic of interest published in high-reputation journals (Q1).
In Administration and has become a concern of other areas in the social sciences and computer science families.
Meanwhile, of the 939 published 'big data and policy' articles, it had 9,160 citations. Table 4 indicates the 25 most cited publication titles. The   Fig. 8 indicates that the network visualization of the article with a minimum of 5 per article citation has formed the 11 network clusters.

Keyword analysis
This analysis is regarded as co-occurrence, which aims to present a visualization of the network between keywords. Based on Fig.9, it is apparent that the analysis using the 'author keywords' of 939 'big data and policy' articles contains 2,879 keywords, in which this research records a minimum number of occurrences from  Source: Processed by Author from Scopus database

Figure 8. Network Visualization from the Article Entitled 'Big Data and Policy'
Source: Data processed from the VOSviewer 1.6.16 software a keyword with 5 repetition keywords. The result obtains 87 keywords that meet the limitation and form 9 clusters (see Fig.9). Fig. 9, it is apparent that the smart city research topic has the largest nodes in cluster 1 (in red), indicating that the topic is predominantly discussed around 2019 (see Fig.10). In addition, Source: Data processed from the VOSviewer 1.6.16 software the 'artificial intelligence' nodes in cluster 4 (in yellow) are larger and are discussed in scientific journals in mid-2019 (see Fig.10). Fig.11 indicates that there are several research topics in the 'big data and policy' area that are rarely discussed (see Fig.11), such as: 'blockchain in big Further, the topics that are most discussed under 'big data and policy' include smart cities, open data, privacy, artificial intelligence, machine learning, data science, social media, big data analytics, and security topics in big data and policy.