Emerging technologies in the renewable energy sector: A comparison of expert review with a text mining software. (March 2020)

• Record Type: Journal Article
• Authors: Moro, Alberto; Joanny, Geraldine; Moretti, Christian
• Is Part Of: Futures. Volume 117(2020)
• Publication Date: 2020-03
• Subjects: Bibliometrics -- Text mining -- Emerging technologies -- Renewable energy -- Quantitative and qualitative -- Policy support; Economic forecasting -- Periodicals; Technological forecasting -- Periodicals; Economic policy -- Periodicals; Prévision économique -- Périodiques; Prévision technologique -- Périodiques; Politique économique -- Périodiques; Economic forecasting; Economic policy; Technological forecasting; Periodicals; Electronic journals; 330.0112
• DOI: 10.1016/j.futures.2020.102511
• Languages: English
• ISSNs: 0016-3287

Abstract

Graphical abstract: Average success of the text mining software in extracting and high ranking (in the top 300) keywords representing the emerging technologies identified by the experts in four renewable energy sectors. Highlights: Expert reviews identified emerging technologies for photovoltaics, wind, ocean energy, and hydropower sectors. Text mining software TIM extracted relevant keywords for the same sectors. The top 300 ranked keywords by TIM are matching those identified by experts with a probability varying from 65 % to 25 %. Technologies with a well-established and univocal jargon are prevailing. Keyword retrieval performances from the TIM and VOSviewer are compared. Abstract: This paper compares the results from quantitative text mining to qualitative expert reviews to identify emerging technologies in the fields of solar photovoltaics (PV), wind power, ocean and tidal energy, hydropower. The text mining analysis is based on the software "Tools for Innovation Monitoring" (TIM). The TIM software extracts a set of relevant keywords from a corpus of pertinent scientific publications. TIM outputs are compared to those extracted by the software VOSviewer, showing agreement. The top 300 ranked keywords are the optimum trade-off between retrieved technologies and analyst efforts. The emerging technologies identified by the experts can be retrieved in the top 300 keywords with a probability of 65 %, 25 %, depending on the technology sector and the algorithm adopted. The more salient keywords tend to correspond to technologies with an established and univocal jargon such as: "dye sensitised solar cells" or "vertical axis wind turbines". Two methods are here used and compared: the frequency of author keywords and the term frequency-inverse document frequency (TF-IDF) algorithm. The comparison of their performances is not showing a general prevalence of one method against the other, but rather a different suitability to different technology sectors.