Big Data in Business Intelligence
Methodology
Data Source
This
study collects literature from datasets of Web of Science Core
Collection (WoSCC) and Scopus. These two databases are involved as the
major citation sources for bibliographic studies.
The search string in WoSCC is set as: TS = ("business intelligen *" OR
"BI") AND ("predict *" OR "forecast" OR "foresee") AND ("big data"). The
search string in Scopus is set as: TS = ("business intelligence" OR
"business intelligent" OR "BI") AND ("predict" OR "prediction" OR
"forecast" OR "foresee") AND ("big data"). Document type = ("article" or
"review"). Time span = (from "1 January 2000" to "7 November 2021").
Analysis Tools
This study utilizes the tool of CiteSpace (5.3.R4, 64-bit) and JRE (1.8) for the literature analysis, accessed on 4 May 2020. This software could be downloaded from the website of https://sourceforge.net/projects/citespace/ (accessed on 18 August 2022), it is generated by Chen etc. According to the "The CiteSpace Manual" released in 2014, CiteSpace I and CiteSpace II are the first and the second version of this tool. The initial publication of CiteSpace I and CiteSpace II from Chen has been cited on Google Scholar 1882 and 4347 times, respectively (The retrieval time is 16 September 2022). CiteSpace is interactive software running based on JRE (1.8) environment, aiming to knowledge extraction, exploring academic achievements, in-depth knowledge graphic visualization, scientific review, and literature quantitative analysis. Developing trend of academic opinions based on time series, contributed scholars, institutions, journals, countries, and discipline subjects are able to be identified and analyzed by using this software, which has widely been utilized in bibliographic studies. CiteSpace generates social networks with nodes and links, which indicates the degree of cooperation between authors, institutions, and countries. The shape of the nodes reflects the influential degree of the author, citation, journal, institution, country, etc. The weight of lines represents the degree of betweenness among nodes. The centrality value reflects the significant degree of nodes, where nodes with a centrality ≥0.1 are regarded as the key nodes.
In this study, the keyword mapping, historic trend, and cluster classification are displayed. The major authors (according to the centrality), core journals (according to the centrality of publications), major institutions (according to the number of publications), most influential counties (according to the centrality of publications), most contributed papers (according to the number of publications), key topics (according to the number of publications), and major involved category (according to the number of publications) are identified.