A keyword selection method for mapping technological knowledge in specific sectors through patent data: the case of biofuels sector

Abstract

In this paper, we propose an innovative methodology that aims to solve drawbacks related to how patent data are allocated and organized in international databases. We propose as a case study in the biofuels sector, in order to evaluate the validity of such a method. Starting with a systematic mapping of biofuels production value chain, we have built a comprehensive description of the biofuels technological domain. The resulting list of keywords relies on an iterative selection approach, based on an analysis of recent scientific literature combined with the keyword search tool developed by Scopus. The final patent database, BioPat, has been finalized by a validation procedure with the help of expert interviews, revealing improved accuracy compared with standard International Patents Classification-based codes. Collected information in BioPat allows us to derive more intriguing insights on the characteristics and evolution of technological patterns in the biofuels sector with respect to standard classification methods.

Keywords:

• patents classification
• keyword selection
• innovation patterns
• liquid biofuels
• knowledge complexity

JEL Classification:

• O31
• O33
• Q42

Acknowledgements

The authors thank Alessandra Molina and Alessandro Palma for their contribution to the collection of keywords, Federico Caviggioli for his advice on downloading patents, Vincenzo Alfano, Alessandro Blasi and Carlo Pignatelli for the patent validation process. Financial support from the European Union D.G. Research with the Grant number 266959 in the research project ‘PICK-ME – Policy Incentives for the Creation of Knowledge: Methods and Evidence’ and the Italian Ministry of Education, University and Research (Scientific Research Program of National Relevance 2010 on ‘Climate change in the Mediterranean area: scenarios, economic impacts, mitigation policies and technological innovation’) is also gratefully acknowledged. The usual disclaimers apply.

Notes

1. Nonetheless, very few contributions have analysed the biofuels sector and its technological development in a comprehensive approach (Costantini and Crespi, 2013; Karmarkar-Deshmukh and Pray, 2009).

2. According to the 2011 version of the IPC guide:

the Classification, being a means for obtaining an internationally uniform classification of patent documents, has, as its primary purpose, the establishment of an effective search tool for the retrieval of patent documents by intellectual property offices and other users, in order to establish the novelty and evaluate the inventive step or non-obviousness (including the assessment of technical advance and useful results or utility) of technical disclosures in patent applications. (WIPO, 2011, p. 1)

3. As an example, IPC contains a whole section for weapons. When interested in the investigation of innovation in the weapons industry, researchers should refer to the IPC code F41. On the contrary, IPC does not contain an ad hoc code for the cosmetic industry (that would probably fall under several categories in the chemistry section).

4. This is the case for sectors that are sufficiently wide to cover an entire section of the IPC such as information and communication technologies, biotechnologies or nanotechnologies, where ad hoc selection tools are also available in search engines such as OECD Patstat. Patents related to these industries are scattered in more than one section but could be easily identifiable. Another example is given by the fertilizer industry. Fertilizers belong to class C05, an ad hoc category in the chemistry section, but fertilizer distributors or fertilizing practices are included in the A01 class, which contains patents related to agricultural activities. Thus, a patent referring to a new fertilizer and its distribution practice would fall under both classes, ideally making a class family that identifies the fertilizer industry.

5. Personal correspondence with WIPO Green Inventory Project staff.

6. The most recent effort in classifying environmental-friendly technologies is given by the Cooperative Patents Classification (CPC), consisting in a new common classification scheme initiated as a joint partnership between USTPO and EPO. The CPC is operative at the two offices by 1 January 2013 and replaced European Classification System (ECLA) at the EPO from that date; it will replace the USPC at the USPTO during a period of transition to 2015. Although CPC seems to be promising since it is more focused on punctual technologies and it has an ad hoc environmental-friendly technologies class, named Y02, it is also true that it is still based on classes mainly, but more importantly it is still under validation and it is a reliable search tool only for recent patents.

7. Conventional biofuels are bioalcohols (especially ethanol) and biodiesel derived from food crops rich in sugar (such as sugar cane and sugar beet) or starch (such as wheat and corn) or vegetable oil (such as palm oil, soya bean and rapeseed). They also include products obtained from a chemical combination of biological molecules with fossil molecules, such as ETBE. Advanced biofuels are bioalcohol based on ligno-cellulosic biomass, HVO (Hydrotreated Vegetable Oil), BTL diesel and Bio-SG (bio-synthetic gas), algae-based biofuels, sugar conversion into diesel-type biofuels, artificial photosynthesis reactions (solar-to-fuel) and hydrocarbons obtained from genetically modified organisms (GMO).

8. The full list of keywords used for BioPat is available upon request from the authors.

9. The list of keywords was built on different levels. First, we tested searches with selected keywords in order to verify the response of the TI database to the inputs and exclude the possibility that the keyword list was inadequate. Second, we created sub-groups of terms standing for the same keyword. For example, raw material terms derived from literature scrutiny were usually in English, while some inventors tend to use scientific Latin or Greek names. Scientific names were attributed to the same original (English) keyword. Moreover, the TI search engine also allows symbols to be used as a means of catching variations of the same word as well as plurals. For instance ‘fermented sugar’ was entered as ‘ferment* sugar*’, catching in this way a combination of different words such as ‘fermenting sugars’ or ‘ferment sugar cane’ and so on.

10. For example: Nannochloropsis (an alga) AND ‘renewable *ethanol’ OR ‘green *diesel’ OR *methanol OR *buthanol OR biomethane OR bio-methyl-ether OR ‘Synthet* fuel*’ OR biodiesel OR ‘renewable fuel*’ OR biofuel* OR etc.

11. After that, we verified if the downloads could represent a significant part of the whole universe achieved using only the general keywords. The huge specific outcome obtained by using general keywords strongly reinforces the choice of working with selected specific keywords rather than working on a broader definition of biofuels (Karmarkar-Deshmukh and Pray, 2009) or on IPC codes.

12. Note that 1,293,197 patents refer to the sum of all patents downloaded (general set + specific keyword for each category) for all kind codes. Such methodology allows for several repetitions; in fact, one patent might be downloaded more than once if it contains more than one keyword.

13. To the best of our knowledge, 35 represents the highest number of countries considered so far in an environmental technology field in international scientific contributions. Our sample includes: AE (Arab Emirates), AR (Argentina), AT (Austria), AU (Australia), BE (Belgium), BR (Brazil), CA (Canada), CH (Switzerland), CN (China), DE (Germany), DK (Denmark), ES (Spain), FI (Finland), FR (France), GB (Great Britain), GR (Greece), HK (Hong Kong), ID (Indonesia), IN (India), IT (Italy), JP (Japan), KP (North Korea), KR (South Korea), LU (Luxembourg), MX (Mexico), MY (Malaysia), NL (The Netherlands), NO (Norway), NZ (New Zealand), PT (Portugal), RU (Russia), SE (Sweden), SG (Singapore), TH (Thailand) and US (United States of America).

14. The exact class codes indicated in GI describing the biofuels domain are reported in Table A1 in the Appendix.

15. We acknowledge that EPO has a very high rate of application refusals and patent applications would not properly reflect the production of new knowledge. Nevertheless, patent applications offer a clear picture of the productive innovative effort. Moreover, working with granted patents obliges us to cut the series considering the years required for the patents acceptance process (the lag between application date and grant date), as suggested by EPO online forum (Johnstone, Hascic, and Popp, 2010; Picci, 2010), losing about five years of information. Moreover, patenting procedures are usually quite slow and patents granted in 2010 do not precisely reflect innovation activities in the earlier years, especially since applicants tend to apply for patents in a national office before applying to an international one (De Rassenfosse et al., 2012).

16. The IPC Versions have been modified seven times from September 1968 to 31 December 2005 (labelled Version 1, Version 2, etc.). Since 2006, the IPC has been frequently revised (sometimes more than once per year) and each edition is indicated by the year and month of its entry into force.

17. We computed variation in the number of IPC class group, which is the higher detailed class specification available for the whole period (subgroups did not exist in previous versions of the IPC system). IPC version 7, Version 8 and the annual review for the period 2006–2009 have been considered.

18. Extract from patent EP2245166A1: The composition of the invention comprises, in various embodiments, a dietary supplement, a food product, a pharmaceutical formulation, humanized animal milk, an infant formula, a cosmetic item and a biodiesel fuel, for example. A pharmaceutical formulation can include, but is not limited to, a drug for the treatment of neurodegenerative disease, a retinal disorder, age-related maculopathy, a fertility disorder, particularly regarding sperm or testes, or a skin disorder.

19. So far, the majority of developed countries have focused their policy interventions on specific roots as highlighted in Figure 10, without paying attention to the whole functioning and interconnections within the system. An interesting exception is represented by the case of the Swedish pro-biofuels policy framework, whose success can be traced into its comprehensive approach. In 2011, there were in fact 12 different pro-biofuels policies in Sweden, including incentives for biofuels pumps and import tools (Holmgren, 2012).