Purpose
Academic research into entrepreneurship policy is particularly interesting due to the increasing relevance of the topic and since knowledge about the evolution of themes in this field is still rather limited. The purpose of this paper is to analyse the key concepts, topics, trends and shifts that have shaped the entrepreneurship policy research agenda during the period 1990–2016.
Design/methodology/approach
This paper uses text mining techniques, cluster analysis and complementary bibliographic data to examine the evolution of a corpus of 1,048 academic papers focused on entrepreneurship-related policies and published during the period 1990–2016 in ten relevant journals. In particular, the paper follows a standard text mining workflow: first, as text is unstructured, content requires a set of pre-processing tasks and then a stemming process. Then, the paper examines the most repeated concepts within the corpus, considering the whole period 1990–2016 and also in five-year terms. Finally, the paper conducts a k-means clustering to divide the collection of documents into coherent groups with similar content. The analyses in the paper also include geographical particularities considering three regional sub-corpora, distinguishing those articles authored in the European Union (EU), the USA and South and Eastern Asia, respectively.
Findings
Results of the analysis show that inclusion, employment and regulation-related papers have largely dominated the research in the field, evolving from an initial classical approach to the relationship between entrepreneurship and employment to a wider, multidisciplinary perspective, including the relevance of management, geographies and narrower topics such as agglomeration economics or internationalisation instead of the previous generic sectorial approaches. The text mining analysis also reveals how entrepreneurship policy research has gained increasing attention and has become both more open, with a growing cooperation among researchers from different affiliations, and more sophisticated, with concepts and themes that moved the research agenda forward, closer to the priorities of policy implementation.
Research limitations/implications
The paper identifies main trends and research gaps in the field of entrepreneurship policy providing actionable knowledge by presenting the spectrum of both over-explored and understudied research themes in the field. In practical terms the results of the text mining analysis can be interpreted as a compass to navigate the entrepreneurship policy research agenda.
Practical implications
The paper presents the heterogeneity of topics under research in the field, reinforcing the concept of entrepreneurship as a multidisciplinary and dynamic domain. Therefore, the definition and adoption of a certain policy agenda in entrepreneurship should consider multiple aspects (needs, objectives, stakeholders, expected outputs, etc.) to be comprehensive and aligned with its complexity. In addition, the paper shows how text mining techniques could be used to map the research activity in a particular field, contributing to the challenge of linking research and policy.
Originality/value
The exploratory nature of text mining allows us to obtain new knowledge and reveals hidden patterns from large quantities of documents/text data, representing an opportunity to complement other qualitative reviews. In this sense, the main value of this paper is not to advise on the future configuration of entrepreneurship policy as a research topic, but to unwrap the past by unveiling how key themes of the entrepreneurship policy research agenda have emerged, evolved and/or declined over time as a foundation on which to build further developments.
Application of Artificial Intelligence in Stock Market Forecasting: A Critique, Review, and Research Agenda
