Submitted:
07 June 2026
Posted:
09 June 2026
You are already at the latest version
Abstract
The Economics of Science examines resource allocation (funding, personnel, infrastructure) and management models for research. A key challenge is assessing research quality and effectiveness when expert review is costly and research is increasingly large-scale and interdisciplinary. Bibliometric analysis of large abstract database records can help address this by identifying research structures and priorities. This study focuses on the Economics of Science with two objectives: identifying key research areas through publications in OpenAlex and using the Mt-KaHyPar program to partition a hypergraph of keywords for analysis of the subject matter within these blocks. The study utilized bibliometric records from the OpenAlex abstract database, containing 474 million scholarly works. A query returned exact matches for specific keywords related to the economics of research, covering the period from 2023 to 2026. In total, 5,556 records were identified, with 5,275 being unique; out of these, 5,250 included data in the “keywords” and “concepts” fields, which were used for analysis. This study illustrates the effectiveness of keywords from OpenAlex in categorizing bibliometric records into subject areas. By utilizing the Mt-KaHyPar program to partition a hypergraph formed from these keywords, balanced blocks were created, revealing themes in the “Economics of Science.” Four key themes emerged: 1) “The Institutional and Interdisciplinary Foundations of Knowledge Production”; 2) “Ethics, Knowledge Governance, and Corporate-Academic Dynamics”; 3) “Pharmaceutical Economics, Bio-Industrial Innovation, Drug Development, and Agro-Food Technologies”; 4) “Digitalization of Science and Computational Methods.” Notably, the “Ethics” theme significantly overlaps with the core theme based on co-occurring keywords. The topic “Ethics, Knowledge Governance, and Corporate-Academic Dynamics” is identified as highly relevant for further research in the field of the Economics of Science.
Keywords:
economics of science
; bibliometric analysis
; OpenAlex records
; hypergraph partitioning
; Mt-KaHyPar program
Keywords. economics of science; bibliometric analysis; OpenAlex records; hypergraph partitioning; Mt-KaHyPar program
Introduction
The Economics of Science primarily examines the distribution of resources—funding, human resources, and infrastructure—for scientific research, as well as models for managing scientific activities. It addresses a fundamental problem: how to assess the quality and effectiveness of research when traditional expert reviews are costly and the results of scientific research are becoming increasingly large-scale and interdisciplinary? Bibliometric studies of large volumes of records from abstract databases can assist in solving these problems by identifying the structure and priorities of research in a given field of knowledge.
This brief review examines publications on bibliometrics, scientometrics, and the economics of science. It analyzes the application and critique of bibliometric indicators for evaluating scientific research and funding, as well as how bibliometric mapping can reveal scientific structures; furthermore, it discusses the limitations and systematic errors of these methods. Examples of studies evaluating the effectiveness of the science economy are provided. A detailed examination of systematic errors in scientometrics arising from the selection of different classification schemes for disciplines and document types in the compilation of university rankings is presented in [1]. This article aims to propose a methodology that integrates uncertainty levels when reporting bibliometric performance in professional practice, addressing the significant influence of university rankings on research management and the associated methodological controversies.
Alter et al. (2025) analyze [2] the limitations of traditional bibliometric indices in academic plastic surgery, arguing that metrics like the h-index create systemic disadvantages and do not reflect true research quality. The study advocates for a multi-factor evaluation model considering author position, discipline norms, and impact beyond citation volume.
Attempts are being made to investigate the relationship between reviewers’ evaluations during the peer review stage and the subsequent scientific recognition of an article, as measured by the number of citations it receives. For example, the authors of [3] collected data from the OpenReview platform on papers submitted to the major AI conference ICLR (International Conference on Learning Representations). The comparison revealed not only a correlation between expert evaluations and the subsequent number of citations but also a paradoxical effect: articles that generated the most disagreement and dissent among reviewers (high variability in evaluations) tend to accumulate a higher number of citations in the long term.
The growing number of bibliometric reviews across various scientific disciplines calls for a systematic guide to their critical evaluation. In response to this challenge, Anh-Duc Hoang [4] has developed and implemented the structured VALOR model (Verification, Alignment, Logging, Overview, Reproducibility). It is designed to assess the quality of bibliometric studies that use data from multiple sources simultaneously.
The study conducted by Yuret (2026) in the Journal of Informetrics [5] investigates the relationship between researchers' self-identified top publications and bibliometric metrics across four social science disciplines, drawing upon data from 2,331 researchers. The findings reveal that, while there is a general correlation between bibliometric metrics and researchers' self-assessments, a notable 30% of the respondents' top three publications in fields such as political science and sociology are not indexed in Scopus. This discrepancy is attributed primarily to the inclusion of books and non-English works that do not appear in well-known bibliometric databases. The implications of this study shed light on the limitations of bibliometric indicators in fully capturing the scholarly impact of diverse forms of research outputs within the social sciences.
Despite its limitations, peer review is still favored for evaluating research proposals. The emergence of artificial intelligence (AI) and generative AI (GenAI) presents new possibilities for enhancing peer review processes. The paper [6] discusses an intelligent peer review system that focuses on the effective integration of AI and machine learning (ML) technologies.
The paper [7] analyzes the influence of scientific research levels by comparing the scientific standing of OECD countries with eight other economies through bibliometric indicators assessing research performance. The findings reveal a strong correlation between research performance and national economic competitiveness, alongside a moderate correlation between research performance and the propensity for research spending.
The study [8] examines the impact of National Science Foundation funding on the scientific productivity of 11,537 principal investigators. It reveals that the anticipated effects of research funding on productivity decrease over time, showing no significant influence on the quality or quantity of research in the five years following the receipt of grants.
Brian A. Jacob and Lars Lefgren [9] investigate NIH grant applications from 1980 to 2000, revealing that each NIH grant of around $1.7 million results in an increase of one additional publication over five years. This represents a 7% enhancement in publication output associated with the funding.
Enrico Vanino, Stephen Roper, and Bettina Becker [10] analyze the impact of UK Research Council funding on business growth, finding that firms involved in these funded projects experience a 23-24% faster growth in employment and sales over six years compared to similar firms. This effect remains consistent across various matching methodologies. The additionality differs based on firm characteristics, with the most significant impacts observed in manufacturing, smaller firms, and those with lower productivity before funding.
The study [11] examines the connection between research funding and academic productivity at a Chinese polytechnic university over the period from 2005 to 2019. By employing an instrumental-variable approach, it demonstrates that competitive research funding significantly enhances both the quantity and quality of academic publications. The findings indicate that the number of grants contributes to sustained productivity by encouraging collaboration and continuity in projects, especially among early-career scholars. In contrast, the total funding amount facilitates the execution of complex, high-impact research initiatives.
A primary goal of the funding is to enhance R&D-driven performance for the host economy [12]. However, there is a risk that the parent firm may not utilize results from host-funded R&D, which could lead to minimal public funding benefits for the host economy. To investigate this issue, a panel dataset comprising 24,404 observations of Irish firms over 10 years is analyzed. The study assesses the impact of R&D grants and tax credits on the R&D activities of subsidiaries, while also exploring the relationship between these policy-induced efforts and the subsidiaries' performance within the host country.
Even a brief overview reveals the multifaceted nature and relevance of the topic at hand, but a search using the query (“Economics of Science”) AND (“Bibliometric Study” OR “Bibliometric Analysis”) yielded only a few studies that were thematically related.
For example, in the article [13], a bibliometric analysis of the journal Labour Economics (LE) is conducted, identifying dominant themes such as unemployment, wages, and education, while noting a growing interest in interdisciplinary issues such as gender inequality and health indicators. Research methodology has evolved and now incorporates cutting-edge empirical methods. The results show that leading contributions to this field come from scholars in Europe and North America, particularly from the United States, Germany, and the United Kingdom.
Another study [14] examines digital nomadism as part of global digital mobility, analyzing its social consequences and impact on society. A systematic review of publications from the Scopus and Web of Science databases covering the period from 2001 to 2023 was conducted, resulting in the selection of 1,047 relevant articles from various disciplines, such as computer science, business, and the social sciences.
This study aims to address the underexplored topic of the Economics of Science by achieving two primary objectives. First, it seeks to identify key research areas associated with the Economics of Science based on publications indexed in OpenAlex, an extensive open-access abstract database. Second, the research will employ the Mt-KaHyPar program to create a balanced partition of a hypergraph comprised of keywords, which will enable an analysis of the subject matter conveyed by these keywords within the defined blocks.
No studies were found that focused on the application of the Mt-KaHyPar algorithm to partition a hypergraph consisting of keywords into balanced blocks.
The practical value of a balanced partitioning into thematic categories can be useful for distributing the workload in the context of expert evaluation—for example, when awarding grants—given that expert evaluation is a labor-intensive process.
Materials and Methods
The data for this study consisted of bibliometric records exported from the OpenAlex open-access abstract database (a catalog of 474 million scholarly works).
The query was designed to return exact matches for the search term (works?search.exact=). The following keywords were used in the query: "economy of science"(321 records), "economics of science"(633), "economics of research"(83), "economics of R&D"(26), "economic impact of research"(120), "commercialization of research"(1450), "academic entrepreneurship"(2924). The data used covers the 2023–2026 period and is current as of May 20, 2026.
Of the total 5,556 records, 5,275 were unique; of these, 5,250 contained data in the “keywords” and “concepts” fields and were subsequently used in this study.
Note. Searches related to research funding and the effectiveness of scientific research yield significantly more results than those focused strictly on the Economics of Science. This suggests that data concerning funding and its effective management should be studied separately, as they may overshadow the main focus of this research.
The requests were made via the API (https://api.openalex.org/works?), which enabled the export of the most comprehensive data in JSON format, containing fields including “primary_topic,” “topics,” “keywords,” and “concepts.” If using CSV export files directly from https://openalex.org/, they would only contain the “primary_topic.display_name” field. In this study, the “keywords” field was the sole one used.
The “keywords” and “concepts” fields in OpenAlex are not author-provided keywords; rather, they are algorithmically generated using the BGE M3-Embedding model. For more detailed information, please visit the OpenAlex Help Center at https://ourresearch.gitbook.io/help.openalex.org, where it is stated that "Legacy Concepts uses a system adapted from MAG to assign multiple concepts to works using their titles and abstracts" and keywords "was developed in coordination with CWTS".
OpenAlex and CWTS (the Center for Scientific and Technical Studies at Leiden University) jointly developed a methodology for generating keywords, which were subsequently used in this study.
For research convenience and to manage file size, specific fields—namely "id", "doi", "title", "publication_year", "fwci", "citation_normalized_percentile", "keywords", and "concepts"—were extracted from JSON format records using the JQ utility (https://github.com/jqlang/jq) and subsequently converted to CSV format.
To create balanced blocks of keywords describing the subtopics of bibliometric records on the subject of the Economics of Science, a hypergraph was constructed whose hyperedges consisted of sets of keywords from each of the 5,250 records.
This study utilized only a hypergraph with unweighted hyperedges. At this stage of the research, it was crucial to understand the informative value and rationality of constructing balanced keyword blocks for analyzing the subject matter of bibliometric records collected on the topic of interest. Justifying the choice of hyperedge weights is a separate task. It is worth noting that OpenAlex data in JSON format can be used to suggest weights such as “citation_normalized_percentile” or “fwci.”
According to the specifications for calculating metrics within the OpenAlex project — "citation_normalized_percentile to works is a percentile rank of citations normalized by the number of works in the same year and subfield" and "Field-Weighted Citation Impact (FWCI) is a snowball metric that takes into account differences in publication type, field (specifically subfield), and year of publication to help understand the citation impact of a particular publication".
The partitioning of the obtained hypergraph based on OpenAlex keywords was carried out using the Mt-KaHyPar utility [15] with the following parameters: Mt-KaHyPar -h KeyWordsNo.hgr -k 4 -e 0.03 -o km1 --preset-type highest_quality -w > OUT_KeyWordsNo.txt. Here: KeyWordsNo.hgr is a hypergraph in which keywords are replaced by their numerical IDs, OUT_KeyWordsNo.txt — a text file reflecting the assignment of terms to blocks, k 4 — partitioning the hypergraph into 4 blocks, e 0.03 — the permissible imbalance recommended by the developers, km1 --preset-type highest_quality — recommended parameters for high-quality hypergraph partitioning. Note: While other parameters were used in the partitioning experiments, they are not discussed here to keep the article concise; the parameters recommended by the developers yielded good results.
Since OpenAlex uses a controlled vocabulary to identify keywords, no text preprocessing was required. The creation of a hypergraph in .hgr format and other similar operations were performed using standard SQL queries and are not discussed here.
Results and Discussion
Partitioning a Hypergraph into Blocks
Given that the analyzed dataset contained 5,250 records, the number of hyperedges was 5,250. The number of unique keywords was 4,538, which corresponded to the number of nodes in the hypergraph. Dividing the hypergraph into 4 blocks using the parameters specified in the previous section yielded the following results (based on Mt-KaHyPar logs):
km1 = 1836 (In addition to the “cut” metric, this metric takes into account how many additional parts each hyperedge is “split” into—that is, whether it spans more than two blocks)
cut = 1803 (The number of hyperedges that have been cut (i.e., contain vertices that fall into at least two different blocks)
Imbalance = 0.0299559 — a value very close to the specified limit 0.03
Partitioning Time = 1.347940 s
The minor difference (33) between cut (1803) and km1 (1836) indicates that most hyperedges are cut between the two blocks.
The distribution of keywords across the blocks was as follows:
|block 0| = 1169
|block 1| = 1101
|block 2| = 1168
|block 3| = 1100
The maximum difference in the number of keywords per block was 69 terms.
It is impossible to evaluate a topic based on more than a thousand keywords, so a system for ranking these terms must be proposed.
A simple and straightforward metric is the frequency with which a keyword included in a given block appears in the overall list of keywords.
The evaluation was performed by applying three operations in sequence: “Split delimited values into separate rows” for the “keywords” field in the main table, a Left Join with the keyword column values from one of the blocks, and a “Count” operation on the filtered values.
Table 1 shows the 30 keywords from block_0 with the highest “Count” value.
To reduce subjectivity in assessing the research topics described by terms from each block, 50 keywords with the highest “Count” value were selected for each block. A query was submitted to Google’s AI-powered search engine to identify topics most relevant to these keywords. Additional explanations were provided regarding the origin of the data, the research objectives, and the methods used. The author acknowledges that the answer will depend on the specific AI and how the question is phrased. However, Google leverages a comprehensive training dataset, enabling it to craft answers that reflect this extensive context. This study focused on the interpretation of topics indicated by block keywords rather than comparing various AI systems or questions. Perplexity.ai could be suggested as an alternative AI system.
The response received for this block is:
Recommended title: The Institutional and Interdisciplinary Foundations of Knowledge Production.
Keywords used by the AI to select the topic for this section: business, political_science, computer_science, entrepreneurship, sociology, economics, psychology, engineering, knowledge_management, marketing, public_relations, higher_education, law, economic_growth, engineering_ethics, finance, social_science, social_psychology, industrial_organization, commercialization, sustainability.
Table 2 shows the 30 keywords from block_1 with the highest “Count” value.
The response received for this block is:
Recommended title: Ethics, Knowledge Governance, and Corporate-Academic Dynamics.
It should be noted that the three terms mentioned in the proposed topic title do not appear directly in the table, but there are terms with similar meanings: information_ethics, bioethics, ethos, scientific_misconduct; in terms not included in the table, there are: collaborative_governance, data_governance, shared_governance, democratic_governance, transition_management_(governance), information_governance, social_knowledge, knowledge_flow and the term listed in the table — tacit_knowledge. Corporate-Academic Dynamics does not appear in the keywords for this section, but the following terms are included: new_business_development, business_relationship_management, scholarly_communication, corporate_finance, corporate_law, corporate_venture_capital.
This is the most difficult section to interpret, but the AI generates a possible topic title based on the broader context in which the keywords listed in the table appear. As a personal assessment, I would say that issues of research ethics, especially in the context of corporate finance and the general digitalization and bureaucratization of knowledge management, are becoming particularly relevant. Therefore, the identification of such a topic when breaking down the studied hypergraph into blocks deserves attention and may serve as a reason for a more detailed study of this topic.
Keywords used by the AI to select the topic for this section: ethos, transactional_leadership, history_of_science, shareholder, service_quality, bioethics, scientific_misconduct, information_ethics, rationality, interim, teaching_method, feminism, tacit_knowledge, scholarly_communication, new_business_development
Table 3 shows the 30 keywords from block_2 with the highest “Count” value.
The response received for this block is:
Recommended title: Pharmaceutical Economics, and Bio-Industrial Innovation, Drug Development Agro-Food Technologies
Keywords used by the AI to select the topic for this section: pharmacology, food_science, immunology, bacteria, gene, microbiology, chemical_engineering, neuroscience, drug, clinical_trial, pediatrics, traditional_medicine, drug_development, cancer, biochemical_engineering, anesthesia, physical_medicine_and_rehabilitation, radiation_therapy.
Table 4 shows the 30 keywords from block_3 with the highest “Count” value.
The response received for this block is:
Recommended title: Digitalization of Science, and Computational Methods
Keywords used by the AI to select the topic for this section: artificial_neural_network, carbon_fibers, digital_marketing, robot, database_transaction, scalability, benchmark(surveying), robotics, energy_consumption, analytic_hierarchy_process, knowledge_integration, remote_sensing, green_growth, contrast(vision), quantum, convolutional_neural_network, geotechnical_engineering
Given that no similar studies could be found at the time of writing this article (there is the author’s own preprint [16] in which the kaHyPar algorithm is used—which can be described as an earlier version of Mt-KaHyPar with eight blocks, the number of which was determined by the nature of the query to the abstract database), it is difficult to conduct a well-founded comparison of the obtained results with previous studies. Therefore, only note that the thematic division of the bibliometric records used into blocks turned out to be quite well interpretable. Filtering records based on the occurrence of keywords within a block (Left Join) allows for a transparent return to the records themselves, which is very important, since subject specialists more often work with publications rather than keywords.
Thanks to the collaboration between OpenAlex and CWTS, exported bibliometric records now include not only the “keywords” field but also normalized publication metrics such as “citation_normalized_percentile.” This makes it easy to estimate the average number of citations for records containing keywords from a given block.
The “citation_normalized_percentile” metric for each block is: block_0 → 0.602; block_1 → 0.6439; block_2 → 0.6245; block_3 → 0.6088. The records most strongly associated with block_1 have the highest “citation_normalized_percentile,” which is another reason to pay attention to the topic described by the keywords from this block.
Given that the keywords in block_0 (let’s call this the base block) appear most frequently in the records under study, it makes sense to assess which of the three remaining blocks is most closely related to block_0. One of the simplest measures of relatedness is the number of records in which terms from each of the three blocks overlap with terms from the base block within a single record.
Overlaps between the data from the three blocks and the main block:
block_0 → block_1 = 775 of 780 non-empty entries related to keywords from block_1. The proportion of non-overlapping records: (780-775)/780=0.0064.
block_0 → block_2 = 431 of 444 non-empty entries related to keywords from block_2. The proportion of non-overlapping records: (444-432)/444=0.027.
block_0 → block_3 = 632 of 640 non-empty entries related to keywords from block_3. The proportion of non-overlapping records: (640-632)/640=0.0125.
According to this indicator, block_1 is closer to the base one than the others.
With the exception of the base block, the other blocks are only loosely connected to one another; for example: block_1 → block_3 = 11.
Discussion of the Role of Ethical Norms in Economics of Science
This section provides a brief discussion of the role of ethical norms in science, the relevance of which is suggested by the results of the analysis of bibliometric records from the OpenAlex platform presented above.
It should be noted that none of the queries submitted to the abstract database mentioned ethical norms. Moreover, the queries did not contain terms related to science funding or the effective use of these funds. There are significantly more publications devoted to science funding than those directly related to terms pertaining to the Economics of Science, and they could obscure the topic under consideration, so they were not used.
Next, the focus will be on discussing a specific issue: how the growing importance of science in the economy might affect the research process, and why ethical norms are becoming increasingly important in this context.
The growing role of science in the economy is sparking increasing interest among the business community—particularly large corporations—and government agencies in scientific discoveries and their practical applications [17,18,19]. However, the core values within the academic community, the corporate world, and government and other administrative institutions differ significantly. In science, openness and transparency in research are essential, as they allow for the verification of the reliability and reproducibility of results, as well as their consistency with existing scientific knowledge. For business, profit, market share, the prospect of attracting investment, and competitive advantages are important. Government agencies may be interested in achieving economic and technological sovereignty, the rational use of budget funds, and job creation. At the same time, each group has its own concepts of work efficiency and the methods for achieving it. Or, to put it simply, they think differently [20,21,22].
The digital technology sector—and AI in particular—as well as the biotechnology sector—and pharmacology in particular—are the leading consumers of scientific research findings, a fact partially confirmed by the topics covered in block_3 and block_2, respectively.
The interest of corporations and administrative agencies in utilizing scientific findings leads them to attempt to apply their own criteria for effectiveness when evaluating such findings. These entities have significant leverage over science, primarily through their financial resources and, secondarily, by lobbying for their interests through the promotion of legislative initiatives and other regulatory documents. Another factor influencing this dynamic is the issue of intellectual property, particularly regarding the raw data held by corporations and administrative organizations and the ownership rights to the resulting findings [23,24,25].
The scientific community is interested in investment, access to data and infrastructure—particularly computing power—and direct oversight of the technological processes that utilize their research findings. For corporations and the government, this entails additional costs and the disclosure of information they consider confidential.
To access these resources, research institutions are forced to take into account the interests of the state and corporations, but no one has disproved the adage that “he who pays the piper calls the tune,” and that access to resources requires striking a balance of interests. However, a balance of interests is possible only when there is a balance of power, and academic institutions have significantly less power than corporations and administrative bodies.
For organizations with limited resources, the only way to resist the influence of a dominant power is to adhere to certain ethical standards in conducting scientific research, so as not to become a puppet of more influential entities. Thus, the growing importance of transparency, openness, and interdisciplinary and international cooperation is not a passing trend, but a means of survival for the scientific community and a way to prevent it from becoming a satellite of more influential communities.
Ethical standards are essential for protecting the scientific community from the overpowering influence of material interests, particularly from corporate and administrative bureaucracies. These entities often struggle to grasp complex scientific issues and resort to measurable metrics, like publication counts or citation rates, suggesting that only scientific experts should evaluate outcomes. This reliance leads to an implicit devaluation of scientific knowledge, as publications are viewed differently by bureaucracies—primarily as tools for justifying budget allocations rather than as contributions to the scientific domain. It is crucial to maintain a clear boundary that differentiates these conflicting valuations between the scientific and bureaucratic perspectives [26,27].
The interaction among science, corporations, and government agencies is evolving rapidly, making it difficult to create a balanced legal framework to resolve conflicts. In this context, ethical standards are crucial for defining acceptable boundaries. The prevailing influence of financial interests poses a challenge, as deviating from ethical norms equates to submitting to these interests.
A detailed discussion of the issues raised is beyond the scope of this article, so we will illustrate this point with a single example. The UN’s seventh Sustainable Development Goal was previously formulated as “affordable and clean energy”; it has now been revised to read “Ensure access to affordable, reliable, sustainable, and modern energy for all,” which means that the emphasis on the social aspect of the issue has only intensified [https://sdgs.un.org/goals/goal7]. However, if we compare the results of two queries, we get a “count” of 224,451 for works?search.exact=”clean energy” and a “count” of 11,625 for works?search.exact=”affordable energy,” i.e., a difference of 19.3. Corporations promote their technologies, and governments promote their programs; this influences research funding, while issues of clean energy affordability take a back seat. Scientific publications are becoming part of marketing.
Scientometric measures of publication activity are still in use; however, rather than serving as benchmarks for assessing scientific interest in particular research, they have become formal indicators used to determine the allocation of funding. It is precisely this that has sparked the scientific community’s interest in ethical norms as a means of curbing an excessive focus on financial and bureaucratic metrics.
This issue goes far beyond the stated objectives of this article and requires a separate and detailed analysis. However, in the context of this work, it is important to note that even such a seemingly abstract scientometric approach as the balanced partition of a hypergraph constructed on the basis of keywords from bibliometric records allows for the clear identification of current topics within the field of “Economics of Science”.
Conclusions
This study demonstrates the usefulness of the keywords assigned by OpenAlex to bibliometric records for identifying subject areas.
By partitioning a hypergraph whose hyperedges are constructed from sets of keywords generated by OpenAlex, using the Mt-KaHyPar program, balanced blocks of keywords were obtained, allowing for a clear interpretation of the themes within each block.
Utilizing the occurrence of keywords from each block within the complete sets of keywords in bibliometric records revealed that such a simple metric can be effective for analyzing subject areas.
The analysis identified four key themes in the field of “Economics of Science”: 1) Core Theme (Block 0), provisionally titled “The Institutional and Interdisciplinary Foundations of Knowledge Production”; 2) Theme (Block 1) “Ethics, Knowledge Governance, and Corporate-Academic Dynamics”; 3) Theme (Block 2) “Pharmaceutical Economics, Bio-Industrial Innovation, Drug Development, and Agro-Food Technologies,” 4) Topic (Block 3) “Digitalization of Science and Computational Methods.”
The theme of the “Ethics, Knowledge Governance, and Corporate-Academic Dynamics” block overlaps most significantly with the theme of the core block based on the co-occurrence of keywords from both blocks in the “Keywords” field.
The topic “Ethics, Knowledge Governance, and Corporate-Academic Dynamics” appears to be the most relevant for further research within the research area of the Economics of Science.
Funding
The work was funded by the Ministry of Science and Higher Education of the Russian Federation (State Assignment No. 125021302095-2).
References
- Robinson-Garcia, N.; Vargas-Quesada, B.; Torres-Salinas, D.; Chinchilla-Rodríguez, Z.; Gorraiz, J. Errors of measurement in scientometrics: classification schemes and document types in citation and publication rankings. Scientometrics 2024, 129(10), 6455–6475. [Google Scholar] [CrossRef]
- Alter, N.; Daiem, M.; Pontell, M.E.; Galdyn, I.; Golinko, M.; Perdikis, G.; et al. Limitations of Academic Bibliometric Indices: The Need for More Comprehensive Metrics. Ann. Plast. Surg. 2025, 95(6), 603–606. [Google Scholar] [CrossRef]
- Xie, W.; Jia, P.; Zhang, G.; Wang, X. Are reviewer scores consistent with citations? Scientometrics 2024. [Google Scholar] [CrossRef]
- Hoang, A.D. Evaluating Bibliometrics Reviews: A Practical Guide for Peer Review and Critical Reading. Eval. Rev. 2025, 49(6), 1074–1102. [Google Scholar] [CrossRef]
- Yuret, T. Assessing scientific output through self-evaluation: Evidence from four social science fields. J. Inf. 2026, 20(1), 101757. [Google Scholar] [CrossRef]
- Lathabai, H.H.; Raman, R.; Nedungadi, P. Novelty First Policy-Based Intelligent Review Framework (IRF) for the Evaluation of Research Proposals. Publications 2026, 14(2), 32. [Google Scholar] [CrossRef]
- Abramo, G.; D’Angelo, C.A. The scientific standing of nations and its relationship with economic competitiveness. In PLOS ONE; Haunschild, R., Ed.; 2024; Volume 19, 6. [Google Scholar] [CrossRef]
- Ding, Y.; Moreira, F. Funding and productivity: Does winning grants affect the scientific productivity of recipients? Evidence from the social sciences and economics. Scientometrics 2025, 130(3), 1831–1870. [Google Scholar] [CrossRef]
- Jacob, B.A.; Lefgren, L. The impact of research grant funding on scientific productivity. J. Public Econ. 2011, 95(9–10), 1168–1177. [Google Scholar] [CrossRef]
- Vanino, E.; Roper, S.; Becker, B. Knowledge to money: Assessing the business performance effects of publicly-funded R&D grants. Res. Policy 2019, 48(7), 1714–1737. [Google Scholar] [CrossRef]
- Fan, X.; Qin, H.; Dai, M.W. Grant counts, funding amount, and academic productivity: causal evidence from the NSFC. In Studies in Higher Education; 2026; pp. 1–16. [Google Scholar] [CrossRef]
- Lenihan, H.; Mulligan, K.; Doran, J.; Rammer, C.; Ipinnaiye, O. R&D grants and R&D tax credits to foreign-owned subsidiaries: Does supporting multinational enterprises’ R&D pay off in terms of firm performance improvements for the host economy? J. Technol. Transf. 2024, 49(2), 740–781. [Google Scholar] [CrossRef]
- Saqlain, M.; Merigó, J.M.; Gil-Lafuente, A.M.; Kydland, F.; Amiguet, L. 100 volumes of Labour Economics: A bibliometric retrospective. Labour Econ. 2026, 102902. [Google Scholar] [CrossRef]
- De Sousa, B.V.; Bastida, M.; Vieira, E. Unraveling digital nomadism: a comprehensive bibliometric review of emerging research and trends. J. Glob. Mobil. 2025, 13(1), 107–132. [Google Scholar] [CrossRef]
- Gottesbüren, L.; Heuer, T.; Maas, N.; Sanders, P.; Schlag, S. Scalable High-Quality Hypergraph Partitioning. ACM Trans. Algorithms 2024, 20(1), 1–54. [Google Scholar] [CrossRef]
- Chigarev, B.N. Hypergraph Partitioning for Bibliometric Term Blocking: An Application of the KaHyPar Framework to IEEE Xplore Data. 2026. [Google Scholar] [CrossRef]
- Arora, A.; Belenzon, S.; Kosenko, K.; Suh, J.; Yafeh, Y. The Rise of Scientific Research in Corporate America. Organ. Sci. 2025, 36(4), 1466–1488. [Google Scholar] [CrossRef]
- Gao, L.; Cai, J.; Chen, N.; Zhong, J. The effect of academic scientists on corporate exploratory innovation performance: Knowledge focus versus long-term focus. Res. Int. Bus. Financ. 2025, 77, 102918. [Google Scholar] [CrossRef]
- Zhu, W.; Han, W.; Lu, R.; Lei, J. The impact of enterprises’ scientific capabilities on innovation performance: evidence from an empirical analysis and simulation model. Humanit. Soc. Sci. Commun. 2025, 12(1), 548. [Google Scholar] [CrossRef]
- Leydesdorff, L.; Meyer, M. The Triple Helix of university-industry-government relations. Scientometrics 2003, 58(2), 191–203. [Google Scholar] [CrossRef]
- Lattu, A.; Cai, Y. Institutional logics in the open science practices of university–industry research collaboration. Sci. Public Policy 2023, 50(5), 905–916. [Google Scholar] [CrossRef]
- Ingstrup, M.B.; Aarikka-Stenroos, L.; Adlin, N. When institutional logics meet: Alignment and misalignment in collaboration between academia and practitioners. Ind. Mark. Manag. 2021, 92, 267–276. [Google Scholar] [CrossRef]
- Myers, K. How publicly funded research preserves scientific integrity. Can. J. Sci. Commun. 2025, 1(1), 53–62. [Google Scholar] [CrossRef]
- Fabbri, A.; Lai, A.; Grundy, Q.; Bero, L.A. The Influence of Industry Sponsorship on the Research Agenda: A Scoping Review. Am. J. Public Health 2018, 108(11), e9–e16. [Google Scholar] [CrossRef]
- Sanda, A.; Unimna, F.; Uduak Edet, U. The Commodification of Knowledge: Scopus-Indexed Journal Publication Requirement, An Aberration to Inclusivity in Nigerian Universities- A Critical Examination. SSRN Electron. J. 2025. [Google Scholar] [CrossRef]
- Ali, Sony Emeka; Ndubuisi, Okeke Gerald; Obiorah, Cynthia Amaka R; Aku, Engr. Ugah Theophilus; Nesiama, Omatseyione; Agbakhamen, Engr. Cletus Onyemhese; Okoro, Engr. Ogheneteme Pupu. Ethical Standards in Research: A Professional Imperative. 2025. [Google Scholar] [CrossRef]
- Valkenburg, G.; Dix, G.; Tijdink, J.; De Rijcke, S. Expanding Research Integrity: A Cultural-Practice Perspective. Sci. Eng. Ethics 2021, 27(1), 10. [Google Scholar] [CrossRef]
Table 1.
Top 30 keywords for the zero block.
| KeyWord | Count | KeyWord | Count | KeyWord | Count |
|---|---|---|---|---|---|
| business | 1843 | public_relations | 494 | philosophy | 305 |
| political_science | 1812 | geography | 484 | social_science | 305 |
| computer_science | 1607 | context_(archaeology) | 429 | management | 297 |
| entrepreneurship | 1166 | higher_education | 398 | mathematics | 288 |
| sociology | 1144 | law | 393 | social_psychology | 287 |
| economics | 1096 | medicine | 387 | government_(linguistics) | 282 |
| psychology | 935 | economic_growth | 376 | process_(computing) | 265 |
| engineering | 814 | engineering_ethics | 337 | pedagogy | 260 |
| knowledge_management | 626 | work_(physics) | 332 | industrial_organization | 252 |
| marketing | 592 | finance | 321 | biology | 238 |
Table 2.
Top 30 keywords in the first block.
| KeyWord | Count | KeyWord | Count | KeyWord | Count |
|---|---|---|---|---|---|
| ethos | 7 | comparative_case | 5 | active_learning_(machine_learning) | 4 |
| privilege_(computing) | 7 | causality_(physics) | 5 | teaching_method | 4 |
| philosophy_of_language | 7 | information_asymmetry | 5 | feminism | 4 |
| transactional_leadership | 6 | grant_funding | 5 | enlightenment | 4 |
| history_of_science | 6 | bioethics | 5 | business_transformation | 4 |
| metaphysics | 6 | scientific_misconduct | 5 | business_relationship_management | 4 |
| reification_(marxism) | 6 | information_ethics | 5 | tacit_knowledge | 4 |
| shareholder | 5 | rationality | 5 | terminology | 4 |
| service_quality | 5 | historiography | 5 | scholarly_communication | 4 |
| vulnerability_(computing) | 5 | interim | 4 | new_business_development | 4 |
Table 3.
Top 30 keywords for the second block.
| KeyWord | Count | KeyWord | Count | KeyWord | Count |
|---|---|---|---|---|---|
| pharmacology | 18 | drug | 8 | cancer | 6 |
| food_science | 17 | horticulture | 8 | gut_flora | 6 |
| immunology | 16 | clinical_trial | 8 | biochemical_engineering | 6 |
| organic_chemistry | 15 | pediatrics | 7 | catalysis | 6 |
| bacteria | 14 | traditional_medicine | 7 | pest_analysis | 6 |
| gene | 14 | cohort | 7 | anesthesia | 6 |
| microbiology | 12 | fish_ | 7 | physical_medicine_and_rehabilitation | 6 |
| chemical_engineering | 11 | physical_therapy | 7 | fishery | 6 |
| neuroscience | 10 | metallurgy | 7 | radiation_therapy | 5 |
| endocrinology | 9 | drug_development | 6 | radiology | 5 |
Table 4.
Top 30 keywords for the third block.
| KeyWord | Count | KeyWord | Count | KeyWord | Count |
|---|---|---|---|---|---|
| artificial_neural_network | 10 | benchmark_(surveying) | 5 | contrast_(vision) | 4 |
| carbon_fibers | 7 | robotics | 4 | random_forest | 4 |
| digital_marketing | 6 | energy_consumption | 4 | complement_(music) | 4 |
| robot | 5 | analytic_hierarchy_process | 4 | quantum | 4 |
| database_transaction | 5 | knowledge_integration | 4 | convolutional_neural_network | 4 |
| principal_(computer_security) | 5 | remote_sensing | 4 | wetland | 4 |
| biodiversity | 5 | baseline_(sea) | 4 | nuclear_physics | 4 |
| hydrology_(agriculture) | 5 | water_resource_management | 4 | geotechnical_engineering | 4 |
| scalability | 5 | green_growth | 4 | sketch | 4 |
| feature_(linguistics) | 5 | upstream_(networking) | 4 | entropy_(arrow_of_time) | 4 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
