Patent Universe. Databases: USPTO and NBER. EXTRACTION of GOVERNING RULES of FORMATION First Patent (1975)

Size: px

Start display at page:

Download "Patent Universe. Databases: USPTO and NBER. EXTRACTION of GOVERNING RULES of FORMATION First Patent (1975)"

Gerard McKenzie
5 years ago
Views:

Prediction of Emerging Technologies Using Patent Citation Network Analysis Péter Érdi, Péter Bruck, István Réthy, Jan

College, Kalamazoo, MI 49006, USA Complex Systems and Computational Neuroscience Group, Wigner Research Centre for

University of Technology and Economics, Budapest, Hungary Network and Subscriber Data Management, Nokia Siemens

1 Prediction of Emerging Technologies Using Patent Citation Network Analysis Péter Érdi, Péter Bruck, István Réthy, Jan Tobochnik, Katherine Strandburg, Gábor Csárdi, Kinga Makovi, Péter Volf Center for Complex Systems Studies, Kalamazoo College, Kalamazoo, MI 49006, USA Complex Systems and Computational Neuroscience Group, Wigner Research Centre for Physics, Hungarian Academy of Sciences, Budapest, Hungary Department of Measurement and Information Systems, Budapest University of Technology and Economics, Budapest, Hungary Network and Subscriber Data Management, Nokia Siemens Networks, Budapest, Hungary Department of Sociology, Columbia University, New York, NY, USA New York University School of Law, New York, NY, USA

2 Patent Universe Databases: USPTO and NBER 4 million patents 44 million citations 1790 First Patent (1975) EXTRACTION of GOVERNING RULES of NETWORK DEVELOPMENT and CLUSTER FORMATION

3 Hierarchical Organization Individual: Patents Mesoscopic: 450 classes 120,000 subclasses Macroscopic: 6 categories 36 sub-categories

4 Microscopic Dynamics Attractiveness depends on in-degree and age Double Pareto distribution C i t a t i o n linear preferential attachment Minoxidil - Citation distribution over years Year

5 Mesoscopic Macroscopic Analysis Citation Vector Non-assortative citations Mesoscopic-macroscopic analysis Citation vector: predictor there ARE clusters!

6 Co-citation clustering Emerging technologies can be identified as changes in the co-citation clusters. Hot-spot detection

7 Back-testing: Emergence of new class citation vector similarity matrix clustering splitting process changing clusters

8 The Ranking of Patents: CiteCount

9 Extending the Definition of the Citation Vector the LENGTH of the vector tells, how important well cited a patent is the DIRECTION of the citation vector enables to compare the effect of external patent classes for two patents

10 The Ranking of Patents Page Rank New Rank Number of Citations CiteCount Patent ID. Patent Class

11 Alternative Mechanisms for Technological Evolution Motivation: Palla et al 2007.

12 Alternative Mechanisms for Technological Evolution Example of Birth (Class 435) Class 435 is about Molecular Biology; the topics of the patents are specifically related to genetic engineering. Page New Number of Patent Patent CiteCount Rank Rank Citations ID. Class

13 Mechanism identification - Birth angle-preserving growth ''2012'' vector 500 C l a s s B 200 ''2004'' vector Class A

14 Alternative Mechanisms for Technological Evolution Example of Birth (Class 435) N o n - n o r m a l i z e d In the past two decades no citations from external patent classes reflected a dominant effect to the development of Class 435. It is reasonable to assume, that this class did not emerge from a recombination of known technologies - since genetic engineering is a completely new discipline.

15 Mechanism identification: Recombination The rotation of the Citation Vector can be observed over time: it shows which external classes have the greatest effect on the Citation Vector over time. C l a s s C ''2004'' vector ''2012'' vector Class D

16 Alternative Mechanisms for Technological Evolution Example of Merging (Class 347) Class 347 is about Inkjet and Laser Printers. Page New Number of Patent Patent CiteCount Rank Rank Citations ID. Class

17 Alternative Mechanisms for Technological Evolution Example of Merging (Class 347) N o n - N o r m a l i z e d Class 400 Class Class 358 Class Year

18 Alternative Mechanisms for Technological Evolution Example of Merging (Class 347) 0.25 N o r m a l i z e d Class 358 Class Class 358 Class Year

19 Alternative Mechanisms for Technological Evolution Example of Merging (Class 347) 3.3 micro-meso relationship In Class 347 the Relative Influence of Class 358 versus Class 400 increased in time ( ) CLUSTER Year 2004 Year 2012

20 What s next? On time prediction! Black Swans and Dragon Kings of the Patent universe.

Web of Science Unlock the full potential of research discovery

Web of Science Unlock the full potential of research discovery Hungarian Academy of Sciences, 28 th April 2016 Dr. Klementyna Karlińska-Batres Customer Education Specialist Dr. Klementyna Karlińska- Batres