This section represents the core of our research, containing all the information related to the dataset. In other words, initially some general details are presented, followed by a description of the most significant papers related to the influence of the use of European funds on the business environment, the form of collaboration both at author and country levels, which were the most common clusters of words, and how they came to be grouped based on different clustering techniques.
3.1. Dataset Description
As seen in
Table 3 below, the dataset was built with papers published from 2005 and stretching up to 2024. In order to be able to gather relevant papers for the research, a total of 60 sources were consulted, referring to different specialized publications such as books, journals, etc. From the above-mentioned sources, a total of 74 papers were identified to be analyzed in detail during the course of this work.
In comparison with another bibliometric analysis conducted by
Milošev et al. (
2024), where the impact of European funds was assessed and where a total of 240 papers were identified for the period 1996–2023, it can be stated that, in our case, the difference in the number of papers included in the dataset might be due to the additional condition of the works relating to the occurrence of specific keywords linked to the business environment.
The average number of years since publication is 8.46, which indicates that more papers were written in the last 8–9 years. Taking into account that the interval over which the papers have been extracted is 20 years, it can already be inferred that there has been a greater interest over time. Moreover, on average, the number of citations per paper has increased annually by 5.95%, again indicating a positive trend of increasing impact over time. In terms of citations, each of the 74 papers was cited on average about 4 (4.257) times by other researchers in their studies. Even the total number of 1764 citations is a representative number for the impact brought about by the selected papers within the field studied.
Out of the total number of 194 contributing authors (
Table 4), 18 of them produced papers on their own, while the remaining 176 authors preferred to collaborate to conduct different studies on the influence of the use of European funds in business. On average, there was about 3 (2.68) co-authors per paper, from which it can be concluded that the majority of the papers have 3 authors. Out of the total number of papers, almost 15% (14.86%) were conducted through international collaborations, which may suggest a beneficial global research area in terms of broad knowledge. Also, in terms of the keywords used by the researchers, a total of 260 words are associated with the documents.
In order to gain an even clearer picture of the annual scientific production, in
Figure 3, a graph is presented which shows the growth for the 20 years analyzed of the interest in the field. The vertical axis indicates the number of existing articles, starting from 0 and increasing from 2.5 up to 10, which is the maximum number of articles identified. Even though the scale is created in this way, the number of articles is always a whole number. In 2005, the first article was published and in the following year, 2006, none were identified. Until 2010, the scientific output varied between one and two articles. The year with the highest number of articles is 2020, with a total of 10 publications, followed by 2015, with a total of 9 articles. It can be stated that on average during the analyzed period, about 4 articles were published (3.7).
In the year 2005,
Bacali and Cordos (
2005) set out to investigate entrepreneurial activities with a focus on sources of finance, different constraints, access to European and governmental funds, and the degree of awareness. Moreover, the authors tried to observe what was the impact of different expenditures on market and financial performance, analyzing a sample of 385 entrepreneurs. At the other pole, in 2024
Iacob Pargaru et al. (
2024) wrote a paper aiming to expose the challenges that arise when an entrepreneur wants to implement a project with the help of European funds. As a paper written in Romania, it demonstrated that these funds have helped the economic and social development of the country.
With the aim of validating the trends observed in the descriptive analysis, a Spearman correlation test was applied (
Figure 4) between the year of publication and the number of articles in order to deduce the correlation between them. The results indicate a moderately strong and statistically significant positive correlation (ρ = 0.6,
p = 0.0047), confirming that the upward trend in academic output between 2005 and 2024 was not random. This supports the previously stated idea that scientific interest in the absorption of European funds and the impact on the business environment has grown steadily and significantly over the last two decades.
In order to be able to analyze how many citations the papers selected in the analysis had,
Figure 5 shows a graph illustrating the average citations by year. This indicator is calculated as the ratio between the TC recorded in a given year and the total number of years. As regards the graph, the situation is similar to that in the previous figure. The vertical axis represents the TC, increasing from 0.5 points to 0.5, except that, this time, as it is an average value, it is possible that the value is not a whole number. Thus, as can be seen, in 2005 there was an average of 0.05 citations per year, with only one article published in that year. Gradually, this reaches in 2010 an average of 1.22 and in 2020 even 1.67. As an interpretation, articles published in 2020 were cited on average about twice. The decreases overtime may be due to the time lag between when a paper is actually published and when it starts to register citations.
Applying Kendall’s test (
Figure 6), a positive but weak association was found between the year of publication and the average number of citations per article (τ = 0.18), but this was not statistically significant (
p = 0.289). Thus, the result obtained suggests that, although some peaks in citations can be visually observed in certain years, it cannot be concluded that there is a constant upward trend in the impact measured by citations.
3.2. Sources
The focus so far has been on the papers that make up the database, so we now move on to the publications which have gained exposure (
Figure 7 and
Table 5). The journals identified, totaling 60, registered between seven and one article each. There is only one journal that managed to publish seven articles, “Proceedings of the International Conference on Business Excellence”, followed by “Scientific Papers-Series Management Economic Engineering in Agriculture and Rural Development” with a total of four articles published. After these two, there are five journals with two, articles each: “Economic and Social Development (ESD)”, “Innovation Vision 2020: from Regional Development Sustainability to Global Economic Growth, vol I-VI”, “Local Economy”, “Metalurgia International” and “Vision 2025: Education Excellence and Management of Innovations Through Sustainable Economic Competitive Advantage”.
Figure 7 and
Table 5 illustrate the top 7 journals, with only 3 of the journals with only one article presented, but the real number is higher, with a total of 53 journals with one article each.
Figure 8 represents the output over the period under investigation of the most significant publications, also shown in
Figure 7. It is interesting to note that although “Proceedings of the International Conference on Business Excellence” has the highest number of publications in the field, the articles only began to be published in 2017, even though according to the conference website (
https://bizexcellence.ro/, accessed on 20 March 2025), the conference is in its 19th edition. The seven papers were published as follows: three in 2017, two in 2020, one in 2022, and one in 2024.
By applying the Mann–Kendall test (
Table 6), strong upward trends (between 0.645 and 0.791) were identified, all of which are statistically significant according to the
p-value. Even “Direction” indicates a significant increase in the number of articles on European funds and the business environment during the period analyzed, reinforcing the relevance of the topic in the specialized literature. “Proceedings of the International Conference on Business Excellence” obtained τ = 0.791,
p < 0.001, and “Scientific Papers-Series Management Economic Engineering in Agriculture and Rural Development” obtained τ = 0.785,
p < 0.001, both showing strong upward trends.
Even though an increased interest has been evident in recent years, the journal “Metalurgia International” published in 2009 the first article entitled “Rural Entrepreneurship Financing Opportunities” (
Vintilǎ et al., 2009), followed in 2010 by “The Effect of the Island Syndrome on the Process of European Funding Absorption” (
Neamtu, 2010). Both articles focused on analyzing and promoting both social and economic development through the absorption of European funds in Romania, emphasizing their impact on different sectors.
In 2015, the journal “Innovation Vision 2020: from Regional Development Sustainability to Global Economic Growth, vol I-VI” published two articles in the field (“Peculiarities of Innovation on Romanian SMEs” (
Ceptureanu, 2015) and “Funding Policies for Youth in the Romanian Rural Area” (
Petrescu, 2015)), and the journal “Local Economy” published one (“An Appropriate Tool for Entrepreneurial Learning in SMEs? the Case of the 20 Twenty Leadership Programme” (
Clifton et al., 2015)).
As shown in
Figure 9, there is also an analysis of publications in terms of the H-index. This is an important index because it manages to combine in a single indicator both the number of citations and the number of papers (
Hirsch, 2005).
The journal “Proceedings of the International Conference on Business Excellence” has the highest impact, as would be expected from what has been observed so far, with a total of 30 citations and an H-index of 3. This means that the journal has three papers that have been cited at least three times. Immediately after this journal, with the same H-index equal to 3, is the journal “Scientific Papers-Series Management Economic Engineering in Agriculture and Rural Development”, but this time the TC is 13, which may be due to the lower number of articles published in this field compared to the first journal mentioned. Then follows the journal “Local Economy” with an H-index of 2 and a TC equal to 15. The other journals have an H-index equal to 1, with the TC ranging from 1 to 20.
3.3. Authors
In the following, authors who have left their mark on this field will be studied by researching representative articles. This sub-chapter is essential to understand their impact, allowing for the identification of collaborative networks between researchers, highlighting trends in teamwork and knowledge sharing between institutions. Moreover, the author survey also helps to track the evolution of the research field, indicating emerging directions of interest and new authors who could become influential in the future, stimulating a more connected and performing scientific environment.
To begin with, in
Figure 10, the first 10 authors have been illustrated according to their scientific output. As shown, four authors contributed with two articles each. Regarding Hapenciuc C. V. and Morosan A. A., the two authors collaborated on the papers “The Performance of European Funded Projects” (
Hapenciuc et al., 2012) and “The Effects of Structural Funds Absorption on the Business Environment in the North-Eastern Region of Romania” (
Moroşan et al., 2014), both analyzing the impact of European funds on Romanian enterprises, with a focus on how they are implemented and used by the beneficiaries. In the same situation are the following two authors, Prokop V. and Stejskal J., who collaborated to write the papers entitled “Determinants of Innovation Activities: Public Financing and Cooperation: Case Study of Czech Republic and Hungary” (
Prokop & Stejskal, 2018) and “Fostering Czech Firms? Innovation Performance Through Efficient Cooperation” (
Prokop et al., 2021). The two papers emphasize the importance of collaboration between different entities, focusing on analyzing the impact of public funds, including European funds, on the innovation activities of manufacturing firms in Central and Eastern European countries, in particular the Czech Republic and Hungary. The remaining 190 authors contributed one paper each.
To go into even more detail, for the 10 authors mentioned above (
Figure 10),
Figure 11 was created in order to be able to observe the years of publication as well as the TC. The articles created by Hapenciuc C. V. and Morosan A. A. were published in 2014 and 2012, respectively, with only one of the two managing to register a citation. On the other hand, Prokop V. and Stejskal J. published their two articles in 2018 and 2021, one of them having a total of 17 citations, while the second one recorded 5 citations. The other six authors listed in
Figure 11 each have only one paper published: Andronic A. O., Andronic R. L. and Asalos N. published in 2019 with a maximum TC equal to 11, Andres B. in 2018 with a total of 5 citations, Armijo A. in 2011 without registering any citation, and Bacali L. in 2005 with 1 citation.
Knowing already the authors and their contribution,
Table 7 presents the top 10 affiliations according to the articles published. As can be seen, the “Bucharest University of Economic Studies” has the highest number of published articles, totaling 14 articles. Due to the economic profile and the business-oriented inclination for which this university is known, this result was to be expected. The university in second place managed to register four articles, the “Stefan cel Mare University of Suceava”, and in third place with three articles is the “University of Pardubice”. The rest of the universities that are part of this top group managed to register two articles each.
What is interesting to note is that most of the universities listed in
Table 7 are located in Romania, which indicates a high interest of this country in analyzing the influence of European funds on the prosperity of enterprises.
Figure 12 shows the collaborative network formed among the top 30 authors, according to the work they have conducted. The reason why there are 14 authors in the figure is that there are a number of authors whose research has been conducted individually. We turned on a setting that isolated nodes so they will not be illustrated, and as such they are not included in the picture.
In total, seven clusters were formed, each consisting of two authors. In the following, a description of each cluster and the works conducted by the authors is provided in order to offer more insight into the research conducted in this field.
Cluster #1, in red, is formed by Hapenciuc C. V. and Morosan A. A., authors who have been discussed in the course of our analysis due to their scientific contribution being representative. The first paper of the two was published in 2012, aiming to analyze the complex process of Romania’s integration into the EU, with a special focus on economic efficiency in the context of recurrent economic crises. Moreover, it considered a crucial aspect of this integration, the use of non-reimbursable European funds, which are an opportunity for Romanian businesses. However, the article details the lack of correct information about investors in terms of the obligations of European funds, analyzing the costs and responsibilities assumed and emphasizing the importance of understanding the “social and institutional costs” of the process (
Hapenciuc et al., 2012). The second paper was published in 2014, also in collaboration with Condratov I. The article analyzes the efficiency of the use of European structural funds by companies in the northeast region of Romania, focusing on a comparative analysis of the annual financial statements of the companies that received these funds and those that did not. The results showed no significant differences between the two groups, concluding that the structural funds had a limited impact on businesses in the short term (
Moroşan et al., 2014).
Cluster #2, in blue, is composed of Prokop V. and Stejskal J., who as in the previous case have been mentioned during our analysis. The first paper published by the two of them was in 2018, starting from the premise that innovation plays a very important role for the competitiveness of economies and firms. The article analyzes the effect of innovation, collaboration between industry and academia, and public funding on turnover growth in Hungary and the Czech Republic. It was concluded that both collaboration and European funds are better managed in Hungary, and the article proposed to improve the use of national funds (
Prokop & Stejskal, 2018). In 2021, together with Striteska M. K., the same authors analyzed the influence of national and European subsidies on the innovation performance of firms in the Czech Republic, also focusing on the role of knowledge and cooperation in innovation activities. The research showed that European subsidies had a significant impact on innovation, while national funds had less effect. Cooperation with research institutions was also found to be crucial for successful innovation. The study contributes to the understanding of the effects of public policy on innovation and catching-up in Central and Eastern Europe (
Prokop et al., 2021).
Cluster #3, in green, is composed of Blidisel R. G. and Bunget O. C. Their paper analyzes the use of EU funds in Romania at the SME level, which is only one-quarter of the total number of EU allocations. This delay is considered to be caused by various financial and technical factors, even information asymmetry, which has the effect of limiting SMEs’ access to external funding. The study uses linear regression to demonstrate the relationship between cash flow and investment, using a sample of 15 SMEs that used grant funds between 2010 and 2011 (
Bunget et al., 2013).
Cluster #4, in purple, is formed by Andronic A. O. and Andronic R. L. The authors tried to explain the involvement of Spiru Haret University in the implementation and development of a project designed to support startups in the central part of Romania, as part of a European program where universities are eligible. A total of 371 out of 403 selected participants completed entrepreneurial training courses, supported by academic staff (
Andronic et al., 2019).
Cluster #5, in orange, is composed of Asalos N. and Bostan I. The researchers focused on the influence of European funds on SMEs in the Danube Delta, maintaining a balance between environmental conservation and economic development. Their research analyzed the environmental, economic, and social effects of the funds on SMEs’ competitiveness. It was observed that European funds have made a significant contribution to increasing competitiveness, but emphasized the need to improve access to them in the future (
Bostan et al., 2019).
Cluster #6, in brown, is made up of Baeza M. D. and Bueno P. C. The paper analyzes the Europe 2020 Project Bond Initiative (PBI), which aims to stimulate infrastructure financing by using European funds and attracting institutional investors. Using SWOT analysis, the constraints and attractiveness of this initiative are examined, based on case studies and responses from public consultations. While the initiative has the potential to support infrastructure financing, there are still challenges to be addressed for its effective implementation (
Vassallo et al., 2018).
Cluster #7, in pink, is composed of Buchholz J. and Castane G. G. The research of the two focused on the development of ethical frameworks with a view to emphasizing the responsible use of artificial intelligence (AI) in industry, highlighting the existing gap between principles and their application. The analysis focused on lessons learned from European projects and the importance of risk management in order to bridge the existing gaps in responsible AI implementation in industry (
Vyhmeister et al., 2022).
3.4. Countries
Knowing information about the authors and the universities where they teach, it was considered relevant to identify those countries with the highest scientific productivity (
Figure 13).
Thus, as expected, due to the affiliations observed in the previous sub-chapter (
Table 7), Romania is the country with the highest productivity identified. Starting in 2005, when it registered its first article, it managed to total 42 articles by 2024. In Germany, the first two articles were published in 2007 and by the end of 2024 it had a portfolio of seven articles. In Italy, the first article was published in 2007, reaching a total of nine articles by 2024. Spain published its first three articles in 2010, reaching a total of nine articles in 2024, the same as in Italy. As for Portugal, they published their first article later, in 2011, and by the end of 2024 they had a total of seven articles, as in the case of Germany. Compared to the rest of the countries, the United Kingdom (UK) published its first article much later in 2012, but by the end of 2024, it had a total of 10 articles.
With the later scientific output of different countries examined, we will now discuss which countries formed the top most cited (
Figure 14).
Therefore, the country with the highest number of citations is Romania, with a total number of 80 citations and an average number of citations per article of 2.80. This is followed by Spain with a TC of 56 and an average of 9.30, followed by the UK with 48 TC and an average of 9.60. Hungary, Czech Republic, Ireland, Slovenia, and Italy are the countries with a TC between 30 and 14, with the rest of the countries with a TC below 10.
The most productive countries are identified in
Figure 14. In
Figure 15, they can be visualized using a global collaborative map. The colors represent the following: the higher the productivity of the country, the darker the shade of blue.
Moreover, in order to show the existing collaboration between countries,
Figure 16 shows a collaborative network between them, forming clusters according to the number of collaborations existing for scientific production.
At the global level, a total of five clusters are identified, each of a different size depending on the number of participants. The first cluster, in red, is made up of Belgium, Austria, Portugal, and Spain; the second cluster, in blue, also which is the largest, is made up of Italy, Germany, Ireland, Sweden, Bulgaria, and United States of America (USA). Cluster 3 consists of France, Iceland, Slovenia, and Austria, cluster 4 of the UK and Malaysia, and cluster 5 of Hungary and the Netherlands.
To create different scientific studies, France collaborated with Iceland, Ireland, Slovenia, and Austria, the Germans collaborated with Bulgaria, Ireland, Sweden, and the USA, Italy with Bulgaria and Germany, even managing to publish two articles in collaboration with Germany, and Spain with Belgium and Portugal. What is very interesting to note is that although most of the work was carried out in Romania, this country only carried out work individually or in collaboration with other researchers within the country, without any external collaborations.
3.5. Most Cited Works
This section analyzes the top 10 most cited documents in the field, with the aim of identifying the main research themes within the dataset and highlighting the important topics that have captured the significant interest of the scientific community in the field.
The most cited paper with a TC equal to 39 is the paper by
Romero-Martínez et al. (
2010), published in 2010 and with an average number of citations per year (TCY) of 2.44 (
Table 8). This means that since 2010 and until now, this paper has been cited at least twice per year. The normalized total number of citations (NTC) is equal to two (calculated as the ratio between the TC values, 39 and 19.5, which is the average TC per article in the year 2010). The NTC indicator is calculated by dividing the TC recorded by an article and the average TC recorded by the papers that were published in that year (
Carhuallanqui-Ciocca et al., 2023). In our case, this indicator shows us that the paper in question was rated 2 times better in terms of citations, making a comparison with the values recorded by the rest of the papers. The paper by
Romero-Martínez et al. (
2010) addresses two important issues related to SME financing. First, it aims to provide a description of the various European financing programs available to Spanish firms. Secondly, the paper aimed to carry out an empirical study observing the use of these programs by Spanish SMEs, with innovation as an objective. Furthermore, the paper analyzes the importance of these programs in supporting SMEs’ innovation processes.
In second place is the paper by
Medve-Bálint and Šćepanović (
2020), with a TC of 30, a TCY of 5, and an NTC of 3 (30 divided by 10); the paper was published in 2020. In the paper, the two authors analyze industrial policy in the market economies of Central and Eastern Europe, arguing that the EU has contributed to the recovery of policy space that was lost through integration into a single market. The authors argued that the EU supports this process through transnational industrial policy, which combines cohesion policy with competition policy. Throughout, the paper compares the use of EU funds in Poland and Romania, highlighting that in Poland efficient institutions have succeeded in achieving a balanced distribution of resources, while in Romania weak institutions have led to a high concentration of funds, mainly benefiting large firms.
In third place is the paper by
Zerbinati (
2012), who is the sole author and published this study in 2012, with a TC of 25, a TCY of 1.79, and an NTC of 1.91. The researcher analyzed what are the reasons that lead to the influence of participation in multilevel governance of local governments, using networks at both local and European levels. Moreover, she turned her attention to the European funding process as it relates to local governments, identifying differences between local authorities in Italy and the UK. The article also explores how entrepreneurship theory can help to explain and understand the phenomenon, analyzing the funding process at the European level (
Zerbinati, 2012).
In fourth place is the article by
Pădurean et al. (
2015), published in 2015. The researchers started from the premise that the European Structural Funds are made to support the social and economic harmonization of EU member states. It was observed that, over time, these funds have been adapted to meet the needs of the European single market, aiming to improve infrastructure, stimulate entrepreneurship, and develop human resources. The central objective of the researchers was to assess the effects of accessing those funds under the Regional Operational Program on tourism in the period 2007–2013 (
Pădurean et al., 2015).
In fifth place is the paper by
Turnsek et al. (
2020), published in 2020. Starting from the idea that aquaponics has gained global attention in recent years, with an increasing number of startups in Europe focused on food production using this technology, a network of researchers and entrepreneurs was created between 2014 and 2018 to develop sustainable aquaponics. Two surveys were conducted, one at European level and one for France, highlighting the challenges facing this technology. The results showed that aquaponics is in a phase of “disillusionment”, but progress in the understanding of the processes could lead to exciting developments in the coming years (
Turnsek et al., 2020).
In sixth place are
Diaz-Orueta et al. (
2020), who published their research during 2020. The paper discusses the impact of assistive technologies developed with European support, which have been shown to bridge the digital divide for older people and those with cognitive decline or even dementia. Although the technologies have been integrated into their lives, there are still doubts about whether they meet the needs and wishes of elderly users. Moreover, it discusses important ethical issues such as vulnerability, privacy and autonomy, and provides examples of technologies, highlighting the preferences of older people and the measures needed for their future development (
Diaz-Orueta et al., 2020).
In seventh place is the work of
Prokop et al. (
2021). During our analysis, Prokop V. and Stejskal J. are also mentioned as collaborators on two scientific papers in
Figure 10 and
Figure 11, where they appeared as two of the authors who are part of the top 10 authors according to scientific output over time. Incidentally, they are also observed in
Figure 12 in the collaborative network realized at the author level. As recalled, the work of the researchers focused on the existing European and national subsidies and their impact on innovation, observing a positive impact for the European ones, while a lower impact was observed for the national ones (
Prokop et al., 2021).
In eighth place is the paper by
Dinu et al. (
2020), published in 2020. The researchers, wanted to observe the extent to which European funds were used for rural development and agriculture in Romania in the period 2014–2020. According to previous studies, the authors noticed that Romania had received a considerable amount of money from the EU budget through the help of certain programs. The paper analyzed the number of projects that had gone through the cycle of submission, selection, and contracting until 2019. Moreover, different sub-measures were analyzed, assessing their impact on rural development and correlating the absorption rate of EU funds with economic and demographic indicators (
Dinu et al., 2020).
In ninth place is the study conducted in 2019 by
Bostan et al. (
2019). Similar to the previous situation with the authors Prokop V. and Stejskal J., Bostan I. and Asalos N. are also encountered during our analysis in the collaborative network illustrated in
Figure 12. As mentioned above, the research examined the influence of EU funds on SMEs in the Danube Delta, with a particular focus on the balance between environmental conservation and economic development. The study analyzed the impact of these funds on the competitiveness of SMEs, highlighting that EU funds have significantly contributed to increased competitiveness. However, it emphasized the need to improve access to these funds in the future (
Bostan et al., 2019).
In tenth place is the paper published in 2016 by
Thomas Lane et al. (
2016). This paper explores how local food and drink entrepreneurship, including food tourism initiatives, can contribute to addressing economic and social challenges in rural communities in Wales. The case study investigates the impact of rural and social enterprises in delivering sustainable products with added value for the community. The context of the analysis is based on principles of equitable development, community decision making, and partnerships, within the framework of the One Planet Wales commitment to Sustainable Development and European funding for rural development (
Thomas Lane et al., 2016).
Table 9 was created to summarize the most relevant information related to the first author, the year of publication, the journal in which the paper was or is to be published, its title, and the dataset used in the analysis, especially for the most cited papers at the global level.
3.6. Most Used Words
Throughout this section, the most frequent additional keywords or the keywords at the author level, as well as groups of keywords (bigrams or trigrams), are identified at title level and abstract level, with the aim of understanding the terms that the authors emphasized during their research.
The word cloud in
Figure 17 presents the additional keywords. According to this figure, it can be easily deduced that the following words occurred the most: “growth” and “performance”, both registering five occurrences each. With three occurrences are “impact”, “policy”, and “research-and-development”, and with two occurrences each are “cohesion policy”, “development subsidies”, “enterprises”, “firms”, “governance”, “innovation”, “management”, and “model”. The rest of the words observed in the word cloud each had only one occurrence.
To make a comparison,
Figure 18 shows the most used words when it comes to authors. This time one very prominent word is “European funds” with a total of 13 occurrences, followed by “SMEs” with 7 occurrences and “entrepreneurship” with 6 occurrences. The following words have occur twice: “European union” and “Romania”. With three occurrences each are “agriculture”, “Czech Republic”, “investments”, “rural development”, “structural funds”, and “tourism”; with two occurrences are a total number of 18 words (“business intelligence”, “cohesion policy”, “community”, “competitiveness”, “cooperation”, “e-learning”, “enterprise”, “entrepreneurs”, “erdf”, “innovation”, “policy”, “regional development”, “rural area”, “rural entrepreneurship”, “sme”, “start-up”, “sustainable development”, and “venture capital”); and with one occurrence, there are 21 words.
Given that our theme is related to the absorption of European funds and how they benefit the business environment, it is natural that in both
Figure 17 and
Figure 18 there are terms referring to the following:
European policies or funds: “European funds”, “structural funds”, “erdf”, “cohesion policy”, “development subsidies”, “policy”, “regional development”, or “sustainable development”;
Enterprises: “SMEs”, “entrepreneurship”, “entrepreneurs”, “enterprise”, “start-up”, “business intelligence”, “enterprises”, “firms”, or “venture capital”;
Agriculture and tourism, this being an area in which European funds have been frequently used: “agriculture”, “rural development”, “tourism”, “rural area”, or “rural entrepreneurship”;
Development and innovation: “innovation”, “research-and-development”, “management”, “growth”, “performance”, or “impact”;
Cooperation and governance: “cooperation”, “community”, “governance”, or “model”;
Countries: “Romania” or “Czech Republic”.
In this sub-chapter, terms referring to “European funds/funding/funded” or “European union” have been removed, as the large number of occurrences of these terms is obvious.
At the abstract level, the bigram “structural funds” has been identified with 17 appearances; there are 16 appearances for “rural development”, 13 appearances for “regional development”, 9 appearances for “European level” and “medium enterprises”, 8 appearances for “funded project”, 6 appearances for “cultural institutions”, “innovation performance”, and “projects implemented”, and 5 appearances for the group “Danube Delta”. At the title level, with 6 occurrences is the group “rural development/entrepreneurship”, with 4 occurrences is “structural funds”, with 3 occurrences are “funds absorption” and “cohesion funds”, and with 2 occurrences are “European projects”, “investment projects”, “romanian rural”, “romanian SMEs”, “union funds”, and “absorption effects/process” (
Table 10).
Making a categorization of the groups of words identified at both the title and abstract level, there are those related to European funds and financing—“structural funds”, “union funds”, “funded project”, “absorption effects/process structural funds”, and “investment projects”; those related to rural development—“rural development”, “romanian rural”, and “projects implemented”; those related to entrepreneurship—“entrepreneurship”, “medium enterprises”, “romanian smes”, and “innovation performance”; and those related to institutions and European level “cultural institutions” and “European level”. These categories highlight key aspects about the use of European funds for rural development and entrepreneurship, such as the absorption of funds, the projects financed, and their effects on SMEs and communities in rural areas.
Making a grouping as in the previous cases, taking into account the trigrams obtained at both the title and abstract level, in
Table 11 there are terms referring to development and regional policy—“regional development policy”, “cohesion funds absorption”, “development fund erdf”, “direct lending funds”, and “complementary eu funds”; terms referring to programs and initiatives—“twenty leadership programme” or “activities public financing”; and terms referring to the analysis of proposals and content—“content analysis proposal”.
3.7. Thematic Analysis
In order to generate the thematic map in
Figure 19, a total of 200 words were selected which have a minimum cluster frequency (per thousand documents) equal to 30, with a label size e 0.3 and a number of labels equal to 3. According to the literature, a thematic map has the role of dividing the selected words into centrality and density. Centrality helps us to observe the amount of work on a particular topic, while density indicates the importance that a particular topic has (
Nasir et al., 2020).
Although a total of 17 clusters are presented in
Figure 19, only 12 of them are illustrated. As it is quite difficult to observe the composition of each cluster, as well as the number of occurrences for each word forming it, we decided to create in
Table 12 a much clearer image for each cluster, mentioning also the color and the quadrant to which it belongs.
Using the data in
Figure 19 and
Table 12, it can be seen that in the “Motor Themes”, quadrant, there are five clusters: “cooperation”, “SMEs”, “European”, “business”, and “policies”. In this quadrant are found those words that represent well-developed themes, being representative for the field under study, which is why here the cluster with the most words in its composition can be observed, “European” with 38 words, of which only “European” or “funds” have more than 20 occurrences each. In the “cooperation” cluster, the highest number of occurrences was recorded for the words “cooperation” and “innovation”; in the “SMEs” cluster, there are words like “SMEs”, “collaboration”, or “developing” with a number of occurrences between 2 and 9; in “business”, there are words like “business”, “analysis”, and “portugal”; and in “policies”, there are words like “cross-border” or “eastern”.
In the “Niche Themes”, there are words related to themes that are not so often used in the analysis. Here, two clusters are identified: “financial” and “capital”. Both consist of three words each with a total number of two occurrences. In “Basic Themes”, there are three clusters, “effect”, “social”, and “local”, which indicate words related to themes that are in a process of development such as “startups”. The last quadrant, entitled “Emerging or Declining Themes”, contains two clusters (“enabling” and “implementation”) whose composition is made up of words with two occurrences each.
The totality of the words identified in each cluster plays a very important role in terms of the use and application for EU funds, whether it is the countries that have been identified as the most common in the research in the dataset, the policies applied in each of them, or the techniques applied.
Furthermore, for topic discovery, LDA has been performed on the extracted papers. LDA was employed as a generative probabilistic model commonly used in natural language processing (NLP) to identify latent topics within a text corpus (
Delcea et al., 2024b). To improve topic detection accuracy, the textual data underwent preprocessing steps including lowercasing and punctuation removal. Given that LDA treats tokens like “SMEs” and “small and medium enterprises” as distinct terms, term normalization was also conducted to unify semantically equivalent expressions.
Also, for the LDA model, grid search was used to optimize the
alpha and
eta hyperparameters, aiming to strike a balance between a relatively small number of topics and high topic coherence. As a result, four topics have been extracted, for an asymmetric
alpha and an
eta of 0.01. The LDA model was implemented using the Gensim Python 4.3.3 library (
Řehůřek & Sojka, 2010).
Topic 1 retains 33.3% of the tokens and provides a focus on the Romanian SMEs and funding. The keywords associated with this topic rely on “romania”, “sme”, “investment”, “grant”, “entrepreneur”, “financial”, “impact”, “market”, “structural_fund”, and “absorption”, as depicted in
Figure 20. As the main focus of the works included in this topic, one can mention the role of structural and grant programs, the economic and financial impact of these programs, and the impact of EU funds absorption.
Topic 2 consists of 32% of tokens and feature a series of keywords related to “cooperation”, “resource”, “innovation”, “business”, “social”, “network”, “public”, “eu”, “economy”, “enterprise”, and “country”, putting an emphasize on elements related to innovation and cooperation in the context of the EU economy, the roles of the social and public sector, the key role played by networks and knowledge transfer, and the policies considered at both local and national levels. The most salient words associated with this topic are highlighted in
Figure 21.
Another important topic given the percentage of tokens, namely 23.8%, is Topic 3, which centers around elements associated with rural and regional development. Among the most relevant terms one can mention “program”, “region”, “agricultural”, “measure”, “regional_development”, “rural_development”, “farmer”, “opportunity”, and “investment”, as observed in
Figure 22.
The last topic identified, namely Topic 4, consists of 10.9% of tokens and retains articles dedicated to research in the area of enterprise collaboration in the rural and cross-border regions, as well as tourism seen as a driver of economic development (
Figure 23). Furthermore, topics related to legal and systemic frameworks are also gathered here. The key terms associated with this topic are “rural”, “tool”, “region”, “tourism”, “enterprise”, “sme”, “production”, “system”, “collaboration”, “community”, “law”, and “model”.
Considering all of the above topics, the following coverage can observed: Topic 1—Entrepreneurship and EU Funding Absorption in Romania, Topic 2—Innovation and Cooperative Enterprise in the EU Context, Topic 3—Regional and Rural Development through Agricultural and EU Programs, and Topic 4—Collaborative Rural Enterprises and Cross-Border Development.
As it resulted from the above discussion, as well as from
Figure 20,
Figure 21,
Figure 22 and
Figure 23, it can be noticed that Topic 4 is isolated in the inter-topic distance maps, while Topic 1 and Topic 3 share some common elements related to their focus on Romanian funding and development. Overall, a strong thematic cohesion around entrepreneurship, EU structural funds, regional development, and innovation can be observed.
By comparing the results of the LDA with the thematic map depicted in
Figure 19 and characterized in
Table 12, it can be observed that there is a connection between the identified topics through the two analyses (thematic map and LDA). LDA Topic 1, for example, is highly connected with cluster 10 (in pink) and cluster 5 (in orange) based on the terms overlapping. Furthermore, LDA Topic 2 covers elements related to cluster 7 (in purple) and cluster 11 (in light blue) extracted in the thematic map, which refers to collaborative and innovative enterprise policies, while LDA Topic 3 aligns with the core regional development terms which can be identified in cluster 10 (in pink) and cluster 14 (in yellow) in the thematic map. Lastly, LDA Topic 4 covers elements related to cross-border cooperation and rural enterprises, with keywords common to cluster 9 (in light green), cluster 10 (in pink), and cluster 1 (in red).
Overall, all the LDA topics are connected to the themes in the thematic map, especially to themes associated with the motor themes quadrant.
In addition to the LDA and thematic map analysis, BERTopic was applied, which constructs coherent topic representations using a class-based variant of Term Frequency- Inverse Document Frequency (TF-IDF). This method has demonstrated competitive performance across several topic modeling benchmarks (
Grootendorst, 2022).
The results of BERTopic are highlighted in
Figure 24. As can be seen, four topics have been uncovered, from Topic 0 to Topic 3, as presented in
Figure 25. For each of the four topics, a series of 10 keywords have been extracted; the top 5 keywords associated with each topic can be observed in
Figure 25, as can the number of papers associated with each topic, retained in the
size indicator.
Considering the topics, it can be noticed that Topic 0 is characterized by keywords such as “european”, “project”, “business”, “based”, “projects”, “information”, “data”, “smes”, “collaboration”, and “optimization”. This topic captures collaborative business projects within the European context, focusing on SME collaborations supported by EU funds.
The financial supports mechanisms provided by the EU are highlighted in the works grouped under Topic 1. This topic gathers keywords such as “funds”, “european”, “financial”, “companies”, “structural”, “european funds”, “process”, “investment”, “projects”, and “structural funds”, and puts an emphasis on structural and CF, the investment process, and the financial governance and implementation processes of EU-funded initiatives.
Topic 2 in BERTopic is characterized by words such as “european”, “small”, “smes”, “networks”, “romanian”, “regions”, “development”, “sme”, “romania”, and “innovation”, suggesting a coverage of the themes associated with Romanian and SME development within this context.
Lastly, Topic 3 gathers words such as “rural”, “development”, “rural development”, “european”, “funds”, “romania”, “entrepreneurship”, “areas”, “rural areas”, and “rural entrepreneurship”. As observed, the main emphasis goes on the rural element associated with EU support, with Romanian studies again being highlighted.
Considering the identified BERTopics, the following labels can be associated with each of them: Topic 0—European Business Collaboration and Data-Driven Innovation, Topic 1—Structural Funds and Investment Mechanisms in the EU, Topic 2—SME Networks and Regional Innovation in Romania, and Topic 3—Rural Development and Entrepreneurship under EU Support.
Further comparing the BERTopic results with the previous topics extracted through LDA analysis and theme extraction through thematic mapping, an overlap of topics and themes can be noticed.
For example, BERTopic 0 is connected with LDA Topic 2 and the thematic map cluster 7 dedicated to cooperation. BERTopic 1 matches elements from LDA Topic 1, as well as from thematic map cluster 10 through the European component. BERTopic 3 features elements from LDA Topic 1 and LDA Topic 2, as well as from thematic map clusters 3 and 10 due to the focus on both SMEs and Romania, while BERTopic 4 covers elements from LDA Topic 3 and LDA Topic 4 and clusters 10 and 14.