7. Innovation Systems

The most important insight that has dominated the field of innovation studies in recent decades is the fact that innovation is a collective activity. It takes place within the context of a wider system. This wider system is coined ‘the innovation system’ or ‘the innovation ecosystem’. The success of innovations is to a large extent determined by how the innovation system is build up and how it functions. Analytically, we can distinguish innovation systems among the analytical dimensions of Geographically positioned units of analysis (e.g. countries and regions), economic exchange relations (sectors and firms), and (technological) novelty production. 

Core concepts: Innovation systems

The most important insight that has dominated the field of innovation studies in recent decades is the fact that innovation is a collective activity. It takes place within the context of a wider system. This wider system is coined ‘the innovation system’ or ‘the innovation ecosystem’. The success of innovations is to a large extent determined by how the innovation system is build up and how it functions. The concept of the innovation system stresses that the flow of technology and information among people, enterprises and institutions is key to an innovative process. It stresses the interaction between actors who are needed in order to turn an idea into a successful process, product or service in the marketplace. Many innovation systems are characterized by some flaws that greatly hamper the development and diffusion of innovations. These flaws are often labelled as system failures or system problems. Intelligent and evidence based innovation policy therefore evaluates how innovation systems are functioning, tries to create insight in the system problems and develops policies accordingly. Innovation systems have been categorized into geographical innovation systems (national or regional), sectoral innovation systems (specific socio-economic developments) and technological innovation systems. These three approaches to analyse innovation systems represent three analytical dimensions of the interaction among an ecology of actors (See figure 1). Geographically positioned units of analysis (e.g., firms, institutions), economic exchange relations, and (technological) novelty production cannot be reduced to one another. However, these independent dimensions can be expected to interact to varying extents.

Given these specifications, one can create a model of the three dimensions and their interaction terms as follows:

Figure 1: Three interacting dynamics of Innovation Systems (adapted from Leydesdorff 2010)

The term National System of Innovation originated at the same time in the work of Christopher Freeman and Bengt-Åke Lundval in the late 1980s. The national innovation system approach emphasises the continued importance of national institutions and arrangements; the quality of basic research, workforce skills, systems of corporate governance, the degree of competitive rivalry and local inducement mechanisms, such as abundant raw materials, the price of labor and energy, and persistent patterns of private investment of public procurement all play a determining role in the innovative capacity of a country.

Sectors differ along several dimensions related to technology, production, innovation and demand. And that they differ in the type and degree of change. The empirical evidence (Malerba-Orsenigo, 1996) suggests also the existence of differences across sectors and of similarities across countries in the patterns of innovative activities for a specific sector.


Technological Innovation System is a concept developed within the context of the Innovation System approach focusing on explaining the nature and rate of technological change. A Technological Innovation System can be defined as the set of actors and rules that influence the speed and direction of technological change in a specific technological area. A key article is provided by Hekkert et al., (2007) Functions of Innovation Systems: A new approach for analyzing technological change.

Patents in the Triple Helix context 

Patents are framed in different contexts: in addition to being output to the production system of knowledge, patents also have utility and serve as input to the economic process of innovation (See Leydesdorff, Alkemade, Heimeriks and Hoekstra, 2014 ).  Furthermore, intellectual property in patents is legally regulated, for example, in national patent offices (e.g., Granstrand, 1999). Thus, three environments are relevant to patenting: the context of technological knowledge production, the economic context, and the legal framework of the state. Patents reflect these different contexts; the names and addresses of inventors and assignees can give information on the location of invention, patent classifications, and claims within the patents provide insights in the paths of technological development.

Together, they provide insights in how innovations in the techno-sciences develop as a result of interactions among knowledge developments, market expectations, and local capabilities. This leads the question whether patents and patent maps can provide us with an analytical lens for studying the complex dynamics of technological innovations? Most patent research has focused on the analysis of the economic dimension. The different kinds of information contained in a patent, however, allow us to study also the geography of inventions, the social networks of co-inventors, and the patterns in the global knowledge bases of inventions. The diffusion of a new technology in these different dimensions can be simultaneous, but also delayed or changing directionality.

In order to fully appreciate the complexity of innovation processes and understand its patterns it is necessary to study the patterns that arise in the different dimensions as well as the interactions between these dimensions.

Our focus is on the patenting of the material technologies that are used for photovoltaic cells. Recently (on January 1, 2013), USPTO and EPO introduced a new system of so-called Cooperative Patent Classifications (CPC) that differently from existing patent classifications (such as International Patent Classifications IPC, and its American or European equivalents) can also be indexed with a focus on emerging technologies using specific tags in the new Y-class. Whereas the previous classification systems have grown historically with the institutions and combine patents that cover product and process innovations at different scales, the classification in terms of CPC takes a reflexive turn: technological classes can be formulated from the perspective of hindsight. The new classifications have been backtracked into the existing databases for indexing.

EPO first experimented with the class Y01 (and its subcategories) as an additional tag for nanotechnology patents (Scheu et al., 2006), while USPTO tried to accommodate nanotechnology into a subclass 977 of its existing classification system. “Y01” was subsequently integrated into IPC v8 as class B82. More recently, a new CPC tag for emerging technologies was developed as Y02: “Climate Change Mitigating Technologies” (Veefkind et al., 2012). This latter tag and its subclasses is now operational in both USPTO and PatStat data. The Y-category is used for “general tagging of new technological developments; general tagging of cross-sectional technologies spanning over several sections of the IPC; technical subjects covered by former USPC cross-reference art collections [xracs] and digests”. (See http://worldwide.espacenet.com/classification?locale=en_EP#!/CPC=Y02E10/00; or similarly at http://www.uspto.gov/web/patents/classification/cpc/html/cpc-Y.html )

Existing routines for overlaying patent data to Google Maps (Leydesdorff & Bornmann, 2012) and a map based on aggregated citations among IPC (Leydesdorff, Kushnir, & Rafols, 2012) were first further developed for the purpose of dynamic mapping (in the case of USPTO patents). The resulting routines are available at http://www.leydesdorff.net/software/patentmaps/dynamic. An equivalent system of routines was developed analogously for patents from PatStat; this software and the technical details can be found at http://www.leydesdorff.net/software/patstat. (These webpages also provide instructions about how to generate the various files.) The USPTO interface is accessed online by the routines, while the PatStat data have to be exported from a local installation of the database by using a script (in sql).

Both routines additionally write a file “rao.dbf” which contains Rao-Stirling diversity for both three and four-digit based maps. Rao-Stirling diversity is defined as follows (Rao, 1982; Stirling, 2007):RAO2

where dij is a disparity measure between two classes i and j—the categories are in this case IPC classes at the respective level of specificity—and pi is the proportion of elements assigned to each class i. As the disparity measure, we use (1 – cosine) since the cosine values among all aggregated IPC is used for constructing the base map of three and four digits. The development of the longitudinal development of Rao-Stirling diversity can be visualized after importing the file “rao.dbf” into Excel.

Stirling diversity combines ‘variety’, ‘evenness’ (or balance) and ‘disparity’ (Stirling 2007).

Jaffe (1986) proposed the cosine between the vectors of classifications as a measure of “technological proximity,” but grouped 328 classes in the US classification system (USPC) into 49 categories. Jaffe & Trajtenberg (2002) in their standard work about patent analysis focused on citations at the document level, but they also developed a higher-level classification of “technological fields” on the basis of the 400+ categories of the US classification system of that time at the three-digits level (Hall et al., pp. 414f.). IPC classes have also been organized into “technological fields” by other researchers for the purpose of the mapping with the argument that IPC classes would be “unbalanced” without such clustering (Kay et al., in press; Schoen et al., 2012) or for the purpose of defining innovation policy objectives (e.g., Schoen et al., 2011).

Leydesdorff, Kushnir, and Rafols (2012) argued that cosine-normalization of the citation patterns among classes can solve the problem of size-differences among them, and that one should be cautious about introducing additional indexer effects by developing one’s own classification scheme (Rafols & Leydesdorff, 2009). These authors proposed to develop maps both at the three- and four-digit level of IPC so that the user can make his/her own choice with reference to the objectives of one’s study. For example, the 637 classes at the four-digit level may be more apt for precise studies of the dynamics of technologies, whereas the 129 classes at the three-digit level may be more useful for policy or portfolio analysis.

The IPC-based maps of VOSviewer for the different years can be animated (e.g., in PowerPoint) given the base maps as stable reflections of the aggregate of citation relations among IPC classes of patents between 1975 and 2011, whereas the overlays show the evolution in specific samples. One can animate the webpages of the geo-maps in PowerPoint similarly using the add-on “LiveWeb” at http://skp.mvps.org/liveweb.htm or using JavaScripts such as made by one of us at http://semweb.cs.vu.nl/patents/index.html. An example of such an animation for the 419 USPTO patents in terms of patent classes is provided at http://www.leydesdorff.net/photovoltaic/cuinse2/cuinse2.ppsx.

Final Assignment (total 400 points)

The final assignment consists of three separate presentation; A, B and C.

Part A: In-depth scientometric analysis

In this assignment, you will provide an in-depth analysis of the Innovation Studies section at Utrecht University. As a well-paid consultant, you are asked by the Dean to provide an analysis of the Innovations Studies group. The best presentation will receive 25 bonus points and a bottle of wine. I will make this presentation available for the IS section. The presentation should be clear, concise and self-explanatory.

The Staff of the Copernicus Institute of Sustainable Development, Section Innovation Studies (ISUU) includes all people with Dr. or Prof. Dr before their name. Using ISI Web of Science publication data of these authors, please provide the following analyses, with a clear explanation and interpretation. Make sure to check that only publications of these staff members are included! (Note that all staff have a researcher ID, for example G-1057-2012).

Not all publications that are listed by the staffmembers are included in the ISI Web of Science (some articles are not yet included, others are published in journals that are not included at all in the WoS). That is not a problem. However, make sure NOT to include publications that are not by ISUU scholars.

UPDATE: Please note that SAINT can handle multiple text files as input (e.g. savedrecs1.txt, savedrecs2.txt, etc). Make sure to put all the files in one folder and select them all. SAINT will automatically eliminate duplicates.

  • What are the most popular journals in which ISUU scholars publish?
  • What are the most frequently cited references?
  • What are the most frequently cited journals?
  • What are the most frequently used keywords?
  • What are the most frequently used abstract words?
  • Please provide a visualisation of the co-author network?
  • Please provide a geographical visualization of the co-author network? Where are the most important collaborators located?
  • Please provide a network visualisation of the most important abstract words in the publications of  ISUU?
  • Please provide a network visualisation of the authors based on the similarities in their abstract words?
  • Please download the records of recent publications in one (or more) important Innovation Journal(s) (see previous questions). Combining these records with the publication records of ISUU can you identify adjacent reseach possibilities? (e.g. Topics that are related to existing topics at ISUU?). What promising topics would you suggest for members of the ISUU?
  • What other relevant and interesting insights can you provide?

Part B: Patentometric analysis of locations, sectors and technologies.

In this assignment, you are asked to provide an analysis of emerging technologies with respect to the geographical, sectoral and technological dimensions.

  • In this assignment we use USPTO patent data (CPC class Y GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS). Please discuss briefly the use of patents as indicator of technological activity. Specifically, what are the advantages and limitations of patent data for measuring and modeling geographical specificities, sectoral specificities and technological specificities?
  • Please summarise and discuss the following articles:
  1. Hekkert, M. P., Suurs, R. A. A., Negro, S. O., Kuhlmann, S., & Smits, R. E. H. M. (2007). Functions of innovation systems: a new approach for analysing technological change. . Technological Forecasting and Social Change, 74(4), 413–432.
  2. Lundvall, B.-Å., Johnson, B., Andersen, E. S., & Dalum, B. (2002). National systems of production, innovation and competence building. Research Policy, 31(2), 213–231.
  3. Malerba, F. (2002). Sectoral systems of innovation and production. Research policy, 31, 247–264.
  • What hypotheses can you formulate with respect to the geographical, sectoral and technological patterns of innovation using the insights from the literature? What type of interactions would you expects among those analytical dimensions?
  • Discuss the Stirling framework of diversity for analysing diversity in science, technology and society. What patterns of Stirling diversity would you expect to occur over time in the evolution of a technology?

We use the software provided by Loet Leydesdorff.  The routines that are discussed here, are derived from USPTO patents at http://www.leydesdorff.net/ipcmaps/dynamic.

The routine PatViz enables the user to animate output from the (geo-coded) patent maps produced from USPTO data (at http://www.leydesdorff.net/software/patentmaps/dynamic ) or from PatStat data (at http://www.leydesdorff.net/software/patstat) both locally and at the internet. In this assignment we use USPTO data.

Interactive versions are provided at http://semweb.cs.vu.nl/patents2 or http://www.leydesdorff.net/patviz. The last release of PatViz can be downloaded from https://github.com/Data2Semantics/PatViz/releases for installation at one’s own machine. After unzipping the files, one installs the program and can run it on one’s computer by clicking on the file index.html. (The computer needs to be connected in order to download the Google Maps providing the background.) The resulting files can also be uploaded after replacing the API key of Google Maps in index.html with the one for one’s own machine.

Currently, files can be uploaded with the extensions pat*.txt (e.g., pat1980.txt, pat1981.txt, etc.) as generated by ps_ipcyr.exe for PatStat data; and the files z*.txt generated by usptoyr.exe for USPTO data. Instructions for preparing this data can be found at http://www.leydesdorff.net/software/patstat and http://www.leydesdorff.net/software/patentmaps/dynamic, respectively. Instead of visualizing these files one-by-one and then generate a map for each year consecutively (at http://www.gpsvisualizer.com/map_input?form=data), PatViz reads the time series of files; the program automatically generates the animation using the parameters specified above.

Two demos are also provided using data for CuInSe2 as material technology for PV cells using Y02E10/541 as the Cooperative Patent Classification for the download in USPTO and PatStat, respectively. See for an example at http://www.leydesdorff.net/photovoltaic/patviz .

In addition to the developments in technolocial knowledge production, we can study the geography of inventor addresses and the industy dynamics through patents assignees. However, this assignee data may need cleaning. Some companies in this list may be subsidiaries of a larger mother company. A subsidiary company (daughter company) is a company that is completely or partly owned by another corporation that owns more than half of the subsidiary’s stock, and which normally acting as a holding corporation which at least partly or (when as) a parent corporation, wholly controls the activities and policies of the daughter corporation.

In today’s fast changing environments, companies need to be innovative in order to sustain their market positions and competitive advantages. Companies face considerable pressure to quickly and effectively respond to local market needs, while achieving global efficiency. This has led some companies to recognize the need to leverage innovation that occurs within their subsidiaries to meet global needs. For example, Philips’ subsidiary in Canada created the company’s first color TV; Philips of Australia created the first stereo TV; and Philips of the UK created the first TV with teletext capabilities.

Please check carefully whether you can find subsidiaries in the list of organisations in the paten class under study. Espacenet http://www.epo.org/searching/free/espacenet.html provides information about the patenting profile of these companies and their subsidiaries. Please discuss your observations. Are the companies specialized or diverse? What about the newcomers? What sectors did they come from? Are the patterns of entry and exit in line with your hypotheses?

Part C: Your comments and Suggestions.

Here you can write any remarks regarding the course. Your feedback is not obligatory, but highly appreciated!


Hekkert, Marko P., et al. “Functions of innovation systems: A new approach for analysing technological change.” Technological Forecasting and Social Change 74.4 (2007): 413-432.

Leydesdorff, Loet. “The knowledge?based economy and the triple helix model.” Annual Review of Information Science and Technology 44.1 (2010): 365-417.

Leydesdorff, Loet, et al. “Geographic and Technological Perspectives on” Photovoltaic Cells:” Patents as Instruments for Exploring Innovation Dynamics.”arXiv preprint arXiv:1401.2778 (2014).

Lundvall, B.-Å., Johnson, B., Andersen, E. S., & Dalum, B. (2002). National systems of production, innovation and competence building. Research Policy, 31(2), 213–231.

Malerba, F. (2002). Sectoral systems of innovation and production. Research policy, 31, 247–264.

Stirling, A., 2007. A general framework for analysing diversity in science, technology and society. Journal of the Royal Society Interface, 4, pp.707–719.