Some comments on using of Web of Science Core Collection for bibliometric studies in volume 29 of the World Journal of Gastroenterology

Abstract

The authors employed inappropriate search keywords and strategies in their published bibliometric papers within volume 29 of the World Journal of Gastroenterology. The comment highlights the identified issues, provides evidence, and suggests improved study methodologies. Subsequent results with more appropriate search strategies were presented to address the shortcomings.

Key Words: Web of Science Core Collection; Front page; World Journal of Gastroenterology; Search keywords; Bibliometric

Core Tip: Approximately 1% of the documents retrieved directly from searches fell outside the designated search period specified in each paper. Consequently, it can be inferred that using data directly from the Web of Science Core Collection may be inappropriate due to the potential inaccuracies introduced by this discrepancy. Findings strongly suggest that employing quotation marks (“ ”) for search keywords is essential for ensuring more accurate data analysis results in bibliometric studies. Integrating additional pertinent search keywords based on the original papers can indeed refine and enhance the precision of data analysis results in bibliometric studies. By meticulously curating the list of abbreviation terms and removing irrelevant ones, researchers can enhance the precision and accuracy of their bibliometric studies, resulting in more meaningful and insightful findings.

TO THE EDITOR

In World Journal of Gastroenterology, volume 29, Web of Science Core Collection (WoSCC) was used in six different articles, entitled “Hotspots and frontiers of the relationship between gastric cancer and depression: A bibliometric study”[1], “Global trends and hotspots of treatment for nonalcoholic fatty liver disease: A bibliometric and visualization analysis (2010-2023)”[2], “Intestinal barrier in inflammatory bowel disease: A bibliometric and knowledge-map analysis”[3], “Immunological factors in cirrhosis diseases from a bibliometric point of view”[4], “Research trends on artificial intelligence and endoscopy in digestive diseases: A bibliometric analysis from 1990 to 2022”[5], and “Global trend and future landscape of intestinal microcirculation research from 2000 to 2021: A scientometric study”[6]. Table 1 shows the search strategy in each paper[1-6].

Table 1 Search strategy using the Web of Science Core Collection.

Ref.	Search strategies
Liu et al[1]	TS = (“stomach neoplasm” OR “stomach cancer” OR “stomach tumour stomach carcinoma” OR “gastric carcinoma” OR “gastric carcinoma” OR “gastric cancer” OR “gastric neoplasm” OR “gastric tumour”) AND TS = (“depression” OR “depressive disorder” OR “depressive symptom”) AND PY = 2000-2022 AND DT = (article or review)
Dai et al[2]	TS = ((nonalcoholic fatty liver) OR (non-alcoholic fatty liver disease) OR (NAFLD) OR (MAFLD) OR (metabolic dysfunction-associated fatty liver disease) OR (metabolic associated fatty liver disease)) AND TS = ((therapy) OR (treatment)) AND LA = English AND DT = (article or review) AND PY = 2010-2023 (April 12, 2023)
Zhou et al[3]	TS = (“inflammatory bowel disease” OR “ulcerative colitis” OR “crohn’s disease”) AND TS = (intestin OR gut OR bowel) AND TS = (barrier OR integrity OR permeability) AND LA = English AND DT = article AND PY = 2001-2021
Zhang et al[4]	TS = ((liver cirrhosis OR hepatic cirrhosis OR liver fibrosis) AND (immunologic factor OR immune factor OR immunomodulator OR biological response modifier OR biomodulator)) AND DT = (article or review) AND PY = 2003-2022
Du et al[5]	TS = ((“artificial intelligence” OR “AI” OR “deep learning” OR “machine learning” OR “machine intelligence” OR “computer” OR “computational intelligence” OR “neural network” OR “knowledge acquisition” OR “automatic” OR “automated” OR “feature extraction” OR “image segmentation” OR “neural learning” OR “artificial neural network” OR “data mining” OR “data clustering” OR “big data”) AND (“endoscopy” OR “endoscope”)) AND LA = English AND PY = 1990-2022
Fu et al[6]	TS = ((intestinal microcirculation) OR (intestinal microvascular)) AND LA = English AND DT = (article or review) AND PY = 2000-2021

DT: Document type; LA: Language; PY: Period; TS: Search terms.

COMMENTS ON USING OF WOSCC

The WoSCC was initially designed to aid researchers in finding pertinent literature[7]. However, researchers have increasingly employed it for conducting bibliometric studies, as noted by Ho[7]. It becomes imperative to employ precise data treatment methods when utilizing this database. Particularly, approximately 1% of the documents retrieved directly from searches fell outside the designated search period specified in each paper (Table 2)[1-6]. Consequently, it can be inferred that using data directly from the WoSCC may be inappropriate due to the potential inaccuracies introduced by this discrepancy.

Table 2 Comparison of results using “front page” as a filter.

Ref.	TP searched out	TP in study period	TP with search keywords in their “front page”	TP from the original paper
Liu et al[1]	284 articles and reviews	282 articles and reviews (99.3% of 284 articles)	200 articles and reviews (71% of 282 articles)	153 articles and reviews
Dai et al[2]	13183 English articles and reviews	13133 English articles and reviews (99.6% of 13183 articles)	11280 English articles and reviews (86% of 13133 articles)	10829 English articles and reviews
Zhou et al[3]	4644 English articles	4585 English articles (98.7% of 4644 articles)	3163 English articles (69% of 4585 articles)	4482 English articles
Zhang et al[4]	3074 articles and reviews	3050 articles and reviews (99.2% of 3074 articles)	1604 articles and reviews (53% of 3050 articles)	2873 articles and reviews
Du et al[5]	4252 English documents	4197 English documents (98.7% of 4252 articles)	3820 English documents (91% of 4197 articles)	446 English documents
Fu et al[6]	1352 English articles and reviews	1343 English articles and reviews (99.3% of 1352 articles)	952 English articles and reviews (71% of 1343 articles)	1364 English articles and reviews

TP: Number of publications.

Despite the common convention of using quotation marks (“ ”) to ensure that search keywords are interpreted as phrases, Dai et al[2], Zhang et al[4], and Fu et al[6] did not apply this technique in their studies. For instance, Dai et al[2] employed the following search keywords: ((nonalcoholic fatty liver) OR (non-alcoholic fatty liver disease) OR (NAFLD) OR (MAFLD) OR (metabolic dysfunction-associated fatty liver disease) OR (metabolic associated fatty liver disease)) AND ((therapy) OR (treatment)). This formulation means ((nonalcoholic AND fatty AND liver) OR (non AND alcoholic AND fatty AND liver AND disease) OR (NAFLD) OR (MAFLD) OR (metabolic AND dysfunction AND associated AND fatty AND liver AND disease) OR (metabolic AND associated AND fatty AND liver AND disease)) AND ((therapy) OR (treatment)). It does not mean to search for phrase keywords: ((“nonalcoholic fatty liver”, “non-alcoholic fatty liver disease”, “NAFLD”, “MAFLD”, “metabolic dysfunction-associated fatty liver disease”, or “metabolic associated fatty liver disease”) AND (“therapy” or “treatment”)). Similarly, Zhang et al[4] used the search keywords: ((liver cirrhosis OR hepatic cirrhosis OR liver fibrosis) AND (immunologic factor OR immune factor OR immunomodulator OR biological response modifier OR biomodulator)) which means ((liver AND cirrhosis OR hepatic AND cirrhosis OR liver AND fibrosis) AND (immunologic AND factor OR immune AND factor OR immunomodulator OR biological AND response AND modifier OR biomodulator)), to signify ((“liver cirrhosis” OR “hepatic cirrhosis” OR “liver fibrosis”) AND (“immunologic factor” OR “immune factor” OR “immunomodulator” OR “biological response modifier” OR “biomodulator”)). Lastly, Fu et al[6] utilized the search keywords: ((intestinal microcirculation) OR (intestinal microvascular)) which means ((intestinal AND microcirculation) OR (intestinal AND microvascular)), to represent (“intestinal microcirculation” OR “intestinal microvascular”).

In the article of Dai et al[2], 10356 English articles and reviews were initially retrieved using appropriate search keywords enclosed in quotation marks (“ ”). Among these, 10316 papers (99.6% of 10356 papers) were published between 2010 and 2023. Subsequently, 9571 papers (93% of 10316 papers) were found to contain the search keywords prominently displayed on their “front page”. A significant disparity emerged between the count of 11280 papers retrieved using search keywords without quotation marks (“ ”) and the count of 9571 papers that met the criteria for search keywords with quotation marks (“ ”). For the article by Zhang et al[4], 246 articles and reviews were initially retrieved, with 241 papers (98.0% of 246 papers) published between 2003 and 2022. Among these, 179 papers (74% of 241 papers) prominently displayed the search keywords on their “front page”. A significant disparity became apparent between the count of 1604 papers retrieved using search keywords without quotation marks (“ ”) and the count of 179 papers meeting the criteria for search keywords with quotation marks (“ ”). Regarding the study conducted by Fu et al[6], 342 English articles and reviews were initially searched out, with 339 of these papers (99.1% of 342 papers) being published between 2000 and 2021. Subsequently, during the analysis, it was found that 306 papers (90% of 339 papers) prominently displayed the search keywords on their “front page”. This observation underscores a significant difference between the initial count of 952 papers retrieved using search keywords without quotation marks (“ ”) and the count of 306 papers meeting the criteria for search keywords with quotation marks (“ ”). These findings strongly suggest that employing quotation marks (“ ”) for search keywords is essential for ensuring more accurate data analysis results in bibliometric studies. The discrepancy between the initial retrieval count and the count of papers meeting the specified criteria indicates the potential for misinterpretation of relevant documents when quotation marks are not utilized. Irrelevant papers might be included in a bibliometric study. Therefore, utilizing quotation marks around search keywords is recommended to enhance the precision and reliability of bibliometric analyses.

In the realm of effective bibliometric studies, a prudent approach involves deeming articles without search keywords in their title, abstract, or author keywords as irrelevant to the pursued topic, as advocated by Fu and Ho[8] and emphasized by Ho[9]. In 2011, Ho’s research group pioneered a groundbreaking “front page” approach, incorporating the article title, abstract, and author keywords to enhance bibliometric studies using the SCI-EXPANDED[10,11] and the SSCI[12]. A decade later, advanced search terms, specifically title (TI), abstract (AB), and author keywords (AK) within the SCI-EXPANDED and the SSCI, have been introduced and employed as the primary “front page” methodology in bibliometric studies medical-related topics[13-16]. The application of TI, AB, and AK as the “front page” has also gained traction in recent years across various bibliometric studies[17-19].

The integration of additional pertinent search keywords based on the original papers can indeed refine and enhance the precision of data analysis results in bibliometric studies. For instance, in the study by Liu et al[1], incorporating terms such as “stomach neoplasms”, “stomach cancers”, “stomach carcinoma”, “gastric carcinomas”, “gastric cancers”, “gastric neoplasms”, “depressions”, “depressive disorders”, and “depressive symptoms” would broaden the scope to include relevant literature on gastric cancer and depression. In the research conducted by Dai et al[2], additional search terms, such as “nonalcoholic fatty livers”, “nonalcoholic fatter liver”, “non-alcoholic fatty liver diseases”, “metabolic dysfunction-associated fatty liver diseases”, “metabolically associated fatty liver disease”, and “metabolic-associated fatty liver diseases” would capture a more comprehensive range of studies related to treatment for nonalcoholic fatty liver disease. For the study by Zhou et al[3], incorporating keywords like “inflammatory bowel diseases”, “inflammatory bowels disease”, “ulcerative colitic”, “Crohn disease”, “Crohns disease”, “Crohns diseases”, “barriers”, “permeabilities”, and “guts” would ensure the inclusion of literature relevant to a intestinal barrier in inflammatory bowel disease. In the article by Du et al[5], additional terms such as “computers”, “computerized”, “neural networks”, “automatically”, “automaticity”, “features extraction”, “feature extractions”, “images segmentation”, “artificial neural networks”, “endoscopy”, “endoscopists”, “endoscopist”, “endoscopes”, “endoscopically”, and “endoscopies” would capture studies related to artificial intelligence and endoscopy in digestive diseases. In the study conducted by Fu et al[6], integrating keywords like “intestinal microcirculations”, “intestinal micro circulation”, and “intestinal micro-vascular” would broaden the search to include literature on intestinal microcirculation research. These additional refinements contribute substantially to ensuring a thorough and precise analysis of literature within the respective fields, enhancing the comprehensiveness and accuracy of bibliometric studies.

Indeed, the abbreviation “AI” can have various meanings across different research fields, making it inappropriate as a search keyword, especially in a wide-ranging study that encompasses multiple disciplines like the one conducted by Du et al[5]. The term “AI” could refer to a multitude of concepts such as apnea index, arousal index, adrenal incidentaloma, apoptosis index, aspirin intolerance, allelic imbalance, apnoea index, aortic insufficiency, adrenal insufficiency, apoptotic index, adrenaline injection, angle of incidence, adventitial involvement, absolute indication, apoptotic indices, additional intervention, absorbance index, activity index, ablation index, air index, awake intubation, air insufflation, and airway interventions, among others. Thus, its usage as a search keyword could lead to ambiguity and inaccurate retrieval of relevant literature. By meticulously curating the list of abbreviation terms and removing irrelevant ones, researchers can enhance the precision and accuracy of their bibliometric studies, resulting in more meaningful and insightful findings. Moreover, in the case of Liu et al[1], the utilization of a limited dataset consisting of only 153 articles and reviews published over 22 years for a bibliometric study could be deemed statistically inadequate. Such a small sample size may not adequately represent the breadth and depth of the research field examining the relationship between gastric cancer and depression.

It is indeed concerning that the authors of the six papers published in the 2023 World Journal of Gastroenterology employed inappropriate methods for their bibliometric studies. Such practices can lead to the dissemination of misleading information to journal readers. Authors have a responsibility to ensure the accuracy and integrity of their research methods, reviewers play a crucial role in identifying and highlighting any mistakes or shortcomings in the manuscripts they review, and journal editors must exercise vigilance in assessing the quality and validity of articles submitted for publication.

CONCLUSION

As highlighted by Ho[20], research serves as a pathway to truth, and researchers need to strive for accuracy and innovation in their work. Merely duplicating the same mistakes without making efforts to improve research methods does not contribute positively to the advancement of knowledge. Therefore, it is imperative for all stakeholders involved in the publication process, including authors, reviewers, and editors, to uphold rigorous standards of scientific integrity and to actively work towards improving the quality of research published in academic journals.