Connecting Scientometrics: Dimensions as a Route to Broadening Context for Analyses

Abstract

Modern cloud-based data infrastructures open new vistas for the deployment of scientometric data into the hands of practitioners. These infrastructures lower barriers to entry by making data more available and compute capacity more affordable. In addition, if data are prepared appropriately, with unique identifiers, it is possible to connect many different types of data. Bringing broader world data into the hands of practitioners (policymakers, strategists, and others) who use scientometrics as a tool can extend their capabilities. These ideas are explored through connecting Dimensions and World Bank data on Google BigQuery to study international collaboration between countries of different economic classification.

Keywords: Dimensions; Google BigQuery; World Bank data; bibliometrics; cloud; contextual data; scientometrics; unique identifier.

Figure 1

Four modes of data use in scientometrics. The most general data, often used for the purposes of resource planning and policy is Global Bibliometric Data—it consists of the bibliographic data such the co-authorship graph, the locational information regarding journal reference and the geographic information. It is the least detailed data and often has the least coverage to PID infrastructure of the datasets shown in this diagram. Global Scientometric Data extends the Global Bibliometric Data with the citation graph and may include altmetric data, although this could also be argued to sit in the outer circle depending on whether it is viewed as attention or context. Local/Organizational Scientometric Data is the data contained in local CRIS systems or national repositories. This data is often enhanced by curation by institutional, funder-based or academicians themselves. Any of these three inner circles may be enhanced by subject classification or annotation approaches to add new facets to the data. The outer circle contains global socioeconomic, political, and other contextual data, which is not typically viewed as being part of the scientometric world. In these data we include items that may connect scientometric data through place, time, or other locational information the the broad world.

Figure 2

High-level entity diagram showing the objects in the Google Big Query environment that support Listing 1. Hexagons outlined in red are those contained in the Dimensions dataset. Hexagons outlined in blue are the data from the World Bank Dataset that are used in this article. Hexagons in solid blue are other hexagons in the World Bank Dataset that haven't been used in this article but which could be of significant interest in further studies. The hexagon outlined in gold is the data that appears in both dataset allowing the link and contextualization that we explore in this article. For those wishing to explore this further, Dimensions has released a free COVID-19 dataset on the Google Cloud Marketplace, which has identical structure to the main database.

Listing 1

Listing to produce an author-contribution-weighted summary of collaborations between countries and connect this to World Bank data using the Google Big Query environment.

Figure 3

Global collaboration between high-income, upper-middle-income, lower-middle-income, and low-income countries on publications between 2010 and 2020. Each quadrant in the chord diagram corresponds to 100% of the research output of countries in each of the brackets. Thus, low-income levels of output have not been normalized in proportion to that of lower-middle, upper-middle, and high-income countries. In this view we see the proportion of collaboration between different economic bands (defined in Table 2). A chord diagram visualisation is helpful to understand the interplay between bi-lateral relationships, but it is important to note that the representation is not entirely faithful as it cannot include multi-lateral relationships directly – a higher-dimensions representation would be required to encapsulate these relationships. However, the contributing bi-lateral component of a multi-lateral collaboration do contribute to each arc.

Figure 4

Top 12 high-income countries with the highest non-high-income country collaborations, ordered by amount cumulative proportion of non-high-income collaboration.

Figure 5

Top 12 High-income-country research institutions with highest non-high-income collaborations.

Figure 6

Top 12 high-to-Low-income-country collaborations by field of research ordered by proportion of low-income country collaborations.

Figure 7

Sustainable development goal collaborations from the perspective of low-income countries, ordered by cumulative non-high-income proportion of research.

Figure 8

Funding acknowledged in papers co-authored between high-income countries and lower income groups, ordered by proportion of low-income research.