Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Talk on Value/Quality of patents by patent plaintiff Type

95 views

Published on

Stanford PAE Symposium in May 2017

Published in: Lifestyle
  • Be the first to comment

  • Be the first to like this

Talk on Value/Quality of patents by patent plaintiff Type

  1. 1. Value/Quality of Patents by Patent Plaintiff Type Jay Kesan P, Ph.D., J.D. Professor & Workman Research Scholar University of Illinois College of Law
  2. 2. PAE Patent Quality/Value • Based on litigation behavior, PAEs buy and assert lousy patents • There are differences in patent quality among various PAE types • PAEs take on risk by investing money in patent purchase and enforcement, so they are choosy about what patents they buy and assert • Numerous patent value measures: Received forward citations; Predicted forward citations; Geographical coverage; Backward citations; Number of claims; Length of the first independent claim
  3. 3. Data Sources 1. Stanford NPE Litigation Dataset: Patent Litigation Data, including patent number, litigation filing date, plaintiff types. 2. Thomson Innovation: Patent Value Factors, including patent number, age, length of 1st claim, #claim counts, #geographic coverage, #backward citations, #observed forward citations, #predicted forward citations (JT). 3. NBER: Technology Categories, including patent number, grant date, industry categories.
  4. 4. Data • Size: 4,450 Observations. 78 84 125 149 166 173 205 187 187 200 220 268 622 668 521 597 0 100 200 300 400 500 600 700 800 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Number of Patent Lawsuits/Unique Patent Number by Year; 2000-2015
  5. 5. Data • Patent Plaintiff Types 1. Acquired patents (N=1291) 2. Failed startup (N=23) 3. Individual-inventor-started company (N=646) 4. Product company (N=2186) 5. Individual (N=230) 6. IP subsidiary of product company (N=29) 7. University/Government/Non-profit (N=16) 8. University heritage or tie (N=1) 9. Startup, pre-product (N=8) 10. Corporate heritage (N=11) 11. Corporate-inventor-started company (N=8) 12. Industry consortium (N=1)
  6. 6. Data 1291 23 646 2186 230 29 0 500 1000 1500 2000 2500 Observations across Plaintiff Types
  7. 7. Data 1484 244 158 0 200 400 600 800 1000 1200 1400 1600 IT Life Science Manufacturing Observations across Three Combined Industry Types
  8. 8. Preliminary Methodology • Descriptive Analyses • ANOVA • OLS Regressions
  9. 9. Comparative Statistics Table 2. Statistics for #Claim Counts N Mean Std. Min. Median Max Acquired Patents 1175 26.77 24.39 1 20 354 Failed Startup 18 14.44 7.79 3 14 42 Individual-inventor- started company 569 33.98 38.89 1 23 424 Product company 1949 20.87 27.03 1 17 887 Individual 202 16.13 12.05 1 14 62 IP subsidiary of product company 24 31.33 24.32 7 22.5 116
  10. 10. Comparative Statistics Table 3. Statistics for Geographical Coverage N Mean Std. Min. Median Max Acquired Patents 1291 2.89 3.28 0 1 25 Failed Startup 23 5.43 9.03 0 5 45 Individual-inventor- started company 646 3.80 4.37 0 2 37 Product company 2186 5.12 8.48 0 2 143 Individual 230 2.99 7.98 0 1 111 IP subsidiary of product company 29 5.62 4.69 0 5 15
  11. 11. Comparative Statistics Table 4. Statistics for Backward Citations N Mean Std. Min. Median Max Acquired Patents 1291 27.05 55.98 0 11 715 Failed Startup 23 19.91 10.29 4 24 54 Individual-inventor- started company 646 56.28 99.07 0 16.5 685 Product company 2186 26.35 37.53 0 15 540 Individual 230 21.38 35.44 0 11 202 IP subsidiary of product company 29 33.55 39.23 1 16 137
  12. 12. Comparative Statistics Table 5. Statistics for Observed Forward Citations N Mean Std. Min. Median Max Acquired Patents 1291 120.08 191.75 0 62 2277 Failed Startup 23 63.65 128.21 0 31 622 Individual-inventor- started company 646 124.99 199.14 0 49 1272 Product company 2186 77.50 144.12 0 29 1780 Individual 230 51.04 179.32 0 18.5 2464 IP subsidiary of product company 29 168.86 261.06 0 93 1387
  13. 13. Comparative Statistics Table 6. Statistics for Predicted Forward Citations N Mean Std. Min. Median Max Acquired Patents 1291 144.14 207.35 0 81.33 2374.21 Failed Startup 23 107.63 138.85 0 102.57 679.29 Individual-inventor- started company 646 162.15 236.52 0 73.97 1710.38 Product company 2186 102.99 170.08 0 48.38 2166.81 Individual 230 61.94 180.49 0 29.59 2464.00 IP subsidiary of product company 29 214.09 275.51 0 135.12 1387.00
  14. 14. Comparative Statistics Table 1. Statistics for Length of the 1st Claim N Mean Std. Min. Median Max Acquired Patents 1291 948.56 748.18 0 845 16999 Failed Startup 23 1018.04 698.73 0 1249 3000 Individual-inventor- started company 646 1046.96 1026.89 0 965 14088 Product company 2186 1002.72 1670.94 0 819.5 24374 Individual 230 992.25 905.31 0 831 6407 IP subsidiary of product company 29 998.66 702.04 0 928 2293
  15. 15. Comparative Statistics Table 7. Statistics for Duration Between Patent Grant and Filing of Lawsuit N Mean Std. Min. Median Max Acquired Patents 907 3679.27 2330.94 -1791 4252 8739 Failed Startup 20 321.90 1794.82 -1695 -498 4726 Individual-inventor- started company 425 2550.07 1937.05 -2790 2451 6565 Product company 911 2056.03 2168.45 -4442 1707 7885 Individual 38 1763.11 2307.02 -3570 1768.5 6342 IP subsidiary of product company 18 3542.33 2495.84 -1952 3818 7169
  16. 16. DifferencesamongPatentValueMeasures Table 9. Correlation Between Patent Value Factors Length of the 1st Claim Claim Counts Geo Coverage Backward Citations Observed Forward Citations Predicted Forward Citations Length of the 1st Claim 1 Claim Counts -0.024 1 Geo Coverage 0.0416*** 0.04** 1 Backward Citations 0.034** 0.1592*** 0.0764*** 1 Observed Forward Citations 0.0584*** 0.1369*** 0.0742*** -0.0279* 1 Predicted Forward Citations 0.0648*** 0.1611*** 0.1135*** 0.0629*** 0.9608*** 1 Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
  17. 17. ANOVA for Plaintiff Types Variables F R-squared Adj R- squared N Length of the 1st Claim 0.41 0.00 0.00 4450 Claim Counts 12.27*** 0.03 0.03 3987 Geo Coverage 9.51*** 0.02 0.02 4450 Backward Citations 14.93*** 0.04 0.03 4450 Observed Forward Citations 9.51*** 0.02 0.02 4450 Predicted Forward Citations 9.86*** 0.02 0.02 4450 Note: Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1
  18. 18. 0 1 2 3 4 5 6 7 Adj. Length of the 1st Claim Claim Counts Geo Coverage Backward Citations Observed Forward Citations Predicted Forward Citations Patent Value Measures by Rank for Different Plaintiff Types Acquired Patents Individual-inventor-started company Product company IP subsidiary of product company Failed Startup Individual
  19. 19. Some Thoughts 1. The relationships between patent value measures vary across both technologies/industries and plaintiff types. 2. It is important to look at various patent value measures and explore their heterogeneity across technology/industry types and plaintiff types. 3. The discussed patent value measures are not independent. e.g., Why is the length of the 1st claim positively correlated with other patent value measures? Shorter claim length may not indicate higher patent value. 4. The relationships between predicted forward citations and other patent value factors may not be linear. (Abrams et al. 2013).
  20. 20. Thank You!

×