Measuring Innovation Using Bibliometric Techniques: The Case of Solar Photovoltaic Industry

Abstract

In this paper, we use feature extraction and data analysis techniques for the elucidation of patterns and trends in technological innovation. In studying innovation, we focus on the role of public research institutions (research universities and national laboratories) in the development of new industries. More specifically, we are interested in measuring innovation through research collaborations between these institutions and the private sector. The proposed methods are primarily drawn from the field of bibliometrics – i.e. the analysis of information and trends in the publication of text documents, rather than the contents of these documents. In particular, we seek to explore the relationship between joint publication patterns and trends, R&D funding, technology development choices, and the viability and effectiveness of industry-university collaborations. To focus the discussions and to provide concrete examples of their applicability, this study will have an initial emphasis on the solar photovoltaic (PV) sector in the U.S., though the techniques and general approach devised here will be applicable to a broad range of industries, situations, and locations. Our analysis suggests that interesting information and conclusions can be derived from this line of analysis. The results obtained using our data extraction techniques allow us to identify early technology focus in different areas within solar PV technologies, and to determine potential technology pathways, which is critical for innovation policy in the renewable energy domain.