Visualizing a field of research: A methodology of systematic scientometric reviews.
Chen C, Song M. · 2019
Read Visualizing a field of research: A methodology of systematic scientometric reviews. by Chen C, Song M. (2019) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc6822756.txt.
Introduction
Chen and Song set out a method for mapping a research field, in PLOS ONE. Reviews that chart a whole field depend on first gathering the right set of papers. Miss key works and the map misleads. Search terms alone often leave gaps. The authors propose building the set by following citations instead.
Their method expands outward from seed papers through generations of citing works. Tests grew one seed set to 748 papers in three steps and about 60 times larger in five. The authors show how to judge whether a set fits the field in both focus and coverage. The aim is reviews that others can repeat and check.
Catalog record
| CREATOR | Chen C, Song M. |
|---|---|
| YEAR | 2019 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Visualizing a field of research: A methodology of systematic scientometric reviews By Chaomei Chen, Min Song, Wolfgang Glanzel Published in: PLoS ONE DOI: 10.1371/journal.pone.0223994 ABSTRACT Abstract Systematic scientometric reviews, empowered by computational and visual analytic approaches, offer opportunities to improve the timeliness, accessibility, and reproducibility of studies of the literature of a field of research. On the other hand, effectively and adequately identifying the most representative body of scholarly publications as the basis of subsequent analyses remains a common bottleneck in the current practice. What can we do to reduce the risk of missing something potentially significant? How can we compare different search strategies in terms of the relevance and specificity of topical areas covered? In this study, we introduce a flexible and generic methodology based on a significant extension of the general conceptual framework of citation indexing for delineating the literature of a research field. The method, through cascading citation expansion, provides a practical connection between studies of science from local and global perspectives. We demonstrate an application of the methodology to the research of literature-based discovery (LBD) and compare five datasets constructed based on three use scenarios and corresponding retrieval strategies, namely a query-based lexical search (one dataset), forward expansions starting from a groundbreaking article of LBD (two datasets), and backward expansions starting from a recently published review article by a prominent expert in LBD (two datasets). We particularly discuss the relevance of areas captured by expansion processes with reference to the query-based scientometric visualization. The method used in this study for comparing bibliometric datasets is applicable to comparative studies of search strategies. INTRODUCTION Systematic reviews play a critical role in scholarly communication [1]. Systematic reviews typically synthesize findings from original research in a field of study, assess the degree of consensus or the lack of it concerning the state of the art in the field, and identify challenges and future directions. For newcomers to a field of study, a timely and comprehensive systematic review can provide a valuable overview of the intellectual landscape and guide new researchers to pursue their research effectively. For experienced and active researchers, systematic reviews can be instrumental in keeping their knowledge of the field up to date, especially when involving areas that are potentially relevant but fall outside the immediate topic of one’s interest. Researchers have also studied scientific literature to uncover potentially significant but currently overlooked connections between disparate bodies of literature, notably, as demonstrated in the research of literature-based discovery (LBD) [2–7]. A fast-growing trend is the increase of systematic reviews conducted with the assistance of science mapping tools [8]. A science mapping tool typically takes a set of bibliographic records of a research field and generates an overview of the underlying knowledge domain, e.g. as with CiteSpace [9, 10] and VOSviewer [11]. For example, systematic reviews facilitated by using CiteSpace include a diverse range of research areas such as regenerative medicine [12], natural disaster research [13], greenhouse gas emission [14], and identifying disruptive innovation and emerging technology [15]. Similarly, VOSviewer was used in reviews of topics such as citizen science [16] and climate change [17]. A scientometric overview of a field of research provides a valuable source of input to conducting systematic reviews, especially in situations when relevant and up-to-date systematic reviews may not be readily available or accessible. The quality of the input data therefore is critical to the overall quality of subsequent analyses and reviews. This practical issue is particularly significant when we need to select subsets of articles from a large pool of available data or when we want to limit the scope of a study to specific disciplines as opposed to open to all disciplines. Considerations concerning the scope of selection have been discussed in the literature in terms of local, global, and hybrid approaches [18]. Global maps of science by definition provide a comprehensive coverage of all scientific disciplines [19], whereas local maps typically focus on selected areas of interest. Hybrid maps may use a global map as a base map and superimpose local maps as overlays [20]. Generating global maps of science requires a substantial array of resources that are not commonly accessible to most of researchers, whereas resources required for generating local maps are more reachable, especially with the increasing accessible science mapping tools. Global maps are valuable as they provide a broad context of a specific research interest. In practice, generating global maps is resource consuming and demanding, for example, requiring direct access to large data sources that are only accessible to a small number of researchers. Consequently, global maps may not be updated as frequently as needed by end users. For example, a significant update of a global model may not be undertaken for 5 years [20]. The underlying structure of a research field is often subject to a variety of changes as the scientific literature grows over time [21, 22]. Therefore, it is reasonable to question whether an existing global model created a few years ago remains a valid representation of the underlying structure for specific analytic tasks in hand, although the answer may differ at different levels of granularity. In contrast, approaches of localism face different challenges. For example, a common challenge for individual researchers of science mapping tools such as CiteSpace [9, 10], VOSviewer [11], and HistCite [23] is to select a representative subset of articles from tens of millions or more of scholarly publications in the Web of Science, Scopus, Dimensions, and/or the Lens. Researchers have used lexical search, citation expansion, and hybrid strategies [24–26]. Complex and sophisticated queries are typically constructed with the input from domain experts and iteratively refined as demonstrated in several in-depths studies [25–27]. We refer such strategies as query-based approaches in this article. The role of relevant and sufficient domain expertise in such approaches is critical, which may lead to a double-edged sword. As we will see in this article, using computational approaches to reduce the cognitive burden from domain experts is a noticeable area of research. Following the principles of LBD to emphasize potentially valuable but currently overlooked connections in scientific literature, it is conceivable that there are undiscovered connections where even domain experts’ input may be limited. Reducing the initial reliance on domain expertise can increase the applicability of query-based strategies. In this article, we propose a flexible computational approach to the construction of a representative dataset of scholarly publications concerning a field of research through iterative citation expansions. Starting from an initial dataset or even a single seed article, the expansion process will automatically expand the initial set by adding articles through citation links in forward, backward, or both directions. This flexibility is particularly valuable in some common scenarios in practice. For example, Swanson’s article on fish oil and Raynaud’s syndrome [6] is widely considered as a groundbreaking one in LBD research. A researcher may want to retrieve subsequently published articles that are connected to the groundbreaking article through potentially lengthy citation paths. Such expansions are valuable to preserve the continuity of research literature identified across an extensive time span. In contrast, a researcher may come across a recently published review article of LBD and would like to find previously published articles that lead to the state of