From Comcast v. Sprint, 2016 U.S. Dist. LEXIS 161623 (opinion by Hon. Jan E. DuBois, ED Pa) on the validity of "forward patent citation analyis" (using later citations to an earlier patent to establish value):
Forward citation analysis is a method of estimating the value of a particular patent based on the number of times the patent is cited by later patents. Comcast Mot. to Exclude, Ex. 19, Expert Report (Corrected) of Alan J. Cox in Connection With Comcast Cable Communication et al. v. Sprint Communications Company L.P. et al. (hereinafter "Cox Report") ¶ 103. Dr. Cox uses forward citation analysis to corroborate his conclusion regarding the reasonable royalty for the '870 patent. Cox Report ¶ 102. Citations are only a comparative indicator of value—"[p]atents in different technologies tend to be cited at different rates, and patents that are older tend to have more citations than patents that were issued more recently." Cox Report ¶ 105. Thus, to ascertain the relative value of the '870 patent, Dr. Cox compiled a pool of patents that are technologically similar to the '870 patent, based on International Patent Classification system labels, and published within six months before and after publication of the '870 patent. Cox Report ¶ 106. Dr. Cox next determined how many times each patent in the resultant pool was cited by later patents. [*18] Cox Report ¶ 107. Using this data, Dr. Cox determined the "percentile ranking" of the '870 patent—that is, "the percentage of categorically similar patents that had less forward citations" than the patent in question. Cox Report ¶ 107.
Comcast first argues that the Court should exclude Dr. Cox's reliance upon forward citation analysis altogether because that method has been "discredited" and therefore fails Daubert's reliability requirement. Comcast Mot. to Exclude at 13-14. In advancing this argument, Comcast relies on a recent trial court decision which it claims rejected the forward citation method as unreliable under Daubert. Comcast Mot. to Exclude at 14; Finjan, Inc. v. Blue Coat Sys., Inc., No. 13-CV-03999-BLF, 2015 U.S. Dist. LEXIS 91528, 2015 WL 4272870 (N.D. Cal. July 14, 2015). But Finjan does not reject forward citation analysis outright—rather, that case recognizes that "a qualitative analysis of asserted patents based upon forward citations may be probative of a reasonable royalty in some instances." Finjan, 2015 U.S. Dist. LEXIS 91528, 2015 WL 4272870 at *8. Comcast also points to a recent paper published by researchers at the University of Pennsylvania in support of its argument. Comcast Mot. to Exclude at 14, Ex. 21 ("Penn Paper").7
[footnote 7: David S. Abrams, Ufuk Akeigit, and Jillian Popadak, Patent Value and Citations: Creative Destruction or Strategic Disruption?, U. Penn. (2013), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2351809. ]
Things get interesting when the court writes:
The authors of the Penn Paper conclude some of the patents with the highest lifetime revenues have fewer citations than patents with median revenues. Penn Paper at 1. However, the forward citation method of analysis has been recognized in the academic literature as reliable since the 1990s.8 Indeed, one meta-analysis of published research on forward citation analysis, which included the Penn Paper, found "forward citation intensity is, in fact, correlated with economic value."9 In short, courts have not rejected forward citation analysis outright. This Court determines that a single academic paper—the Penn Paper—is not sufficient to rebut decades of literature supporting forward citation analysis.
[footnotes 8 and 9: 8 See, e.g., Dietmar Harhoff, Frederic Scherer, and Katrin Vopel, Citation, family size, opposition and the value of patent rights, Research Policy, 1596 (2002); Bronwyn H. Hall, Adam Jaffe, and Manuel Trajtenberg, Market Value and Patent Citations, RAND Journal of Economics 36 (1) (Spring 2005); Manuel Trajtenberg, A Penny for Your Quotes: Patent Citations and the Value of Innovations, RAND Journal of Economics 21 (1) (Spring 1990).9 Adam B. Jaffe and Gaétan de Rassenfosse, Patent Citation Data in Social Science Research: Overview and Best Practices, Nat'l Bureau of Econ. Research Working Paper Series (2016), available at http://www.nber.org/papers/w21868 .]
** On 2 June 2005, there was a post on IPBiz entitled Patent citations, again?
, noting, among other things, a 2003 paper criticizing patent citation:
Of one of CHI's studies, from Lawrence B. Ebert, Say Good-Night, Gracie, Intellectual Property Today (June 2003):
** See also the 2009 IPBiz post
"Trolls on top?" or how not to cite relevant work?
** See also the 2010 IPBiz post
The citations found in patents: do references cited by the applicant matter?
, which contains a reference to a 1990s paper severely criticizing the use of patent citation analysis:
Simmons, Edlyn S.; Lambert, Nancy. "Comparing grapes and watermelons." ChemTech 23 (6), 1993, p. 51-59.
**Note separately the paper titled -- The Doctrine of Equivalents: Hilton Davis and Pioneering Patents -- [Intellectual Property Today, p. 10 (December 1996), available on LEXIS, which includes the text criticizing the patent citation work of Breitzman and Narin ("BN"):
In the March, 1996 issue of The Law Works, Breitzman and Narin ("BN") presented the argument "that very highly cited patents, patents of 'Pioneering Class' citation frequency, may be entitled to a broad interpretation of their claims, as are Pioneer patents." BN cited various studies that "support the notion of the technological importance of the inventions represented by highly cited patents." They noted that "there is a strong statistical association between pioneering patents and high citation frequency, just as there is a strong statistical association between high citation in patents and their technological importance."
In a letter in the November, 1996 issue of The Law Works, BN acknowledged that there were worthless patents that have received a lot of citations, just as there are important patents that have received few citations. Thus, the presence of many citations, by itself, does not prove value, and the absence of citations does not prove lack of value.
Nevertheless, BN find it inexplicable that 64 non-self-citations (out of 165 total cites) for US 4,105,776 don't strike me as extraordinary. My answer is that the number of citations for any patent, in the absence of an analysis of the citations, is insufficient to prove value. A few examples are in order.
In the November, 1996 letter, BN point to an article by Narin and Frame that was featured in the August 11, 1989 issue Science (which they consider "possibly the most prestigious of the thousands of U.S. research journals") n7. According to the ISI database (Dialog 434), this article has been cited 6 times. An article by Ebert in a 1990 issue of Science n8 has been cited 23 times. I would submit that the value of neither article is given by the number of citations.
n7 The Science article pertained to the growth of Japanese science. Narin's work on high citations of Japanese patents was discussed in Business Week (p. 57 of Aug. 9, 1993 issue), in the Jiji Press Ticker (May 28, 1991), Business America (Vol. 110, p. 13, Sept. 25, 1989 issue) and the New York Times (p. A1 of March 7, 1988 issue). He discussed the poor citation of Soviet work in Science, p. 629 (Feb. 8, 1985). Narin recently published an article "Biblometric Performance Measures" in Scientometrics, 1996, 36, 293-310; the abstract notes "The key measure is whether an agency is producing or supporting highly cited papers and patents."
n8 Science, 247, 1469 (1990).
BN referred to a 1986 article by Eugene Garfield, "Do Nobel Prize Winners Write Citation Classics?" For reference, I include herewith the total (first author) citation count for the three Nobel Prize winners in Chemistry in 1996, who won for their work on buckminsterfullerene (C60; "buckyball"): R.E. Smalley (1715), H. W. Kroto (3754), and R. F. Curl (1464). For comparison, I give numbers for past chemistry laureates: L. Pauling (23,187), P. J. Flory (19,173), G. A. Olah (11,795) and H. Taube (1785) and other chemistry professors J. P. Collman (9376), J. I. Brauman (1009), and N. H. Nachtrieb (351) n9.
n9 These prestigious authors presumably publish in primarily prestigious journals. Garfield has noted that only 100 journals (out of 3,400 indexed journals in SCI) account for 44% of cited articles.
In the November, 1996 letter, BN refer to the case I discussed (Gordon and Breach v. AIP) as "one rather old law suit", even though the case is still pending n10. This case was one example of many which illustrated potential issues with citation analysis n11.
n10 Summary judgment matters were disposed of in 859 F. Supp. 1521 (SD NY 1994) and in 905 F. Supp. 169 (SDNY 1995).
n11 We noted the issue of channeling in individual cites (the quote of Maddox and of Daubert) and in journal cites (Energy & Fuels). There are issues of improper cites and nonexistent cites, detailed in L. B. Ebert, FSTI, 15(3), in press, 1997. One can have cites to papers which are scientifically wrong, and perhaps even fraudulent (see C & E News, p. 7 (Nov. 4, 1996)). One notes that in the episode of the television show Law & Order which we cited in the July 1996 issue of The Law Works (p.6), as well as in the episode which aired Nov. 6, 1996, scientists who had all the superficial indicia of success, presumably including many citations, were, after detailed analysis, demonstrated to be villains. A moral of these episodes, which we adopt here, is that one has to look beyond the superficial before drawing conclusions. "Nose counts" of citations alone are just not enough to prove value.
BN made no comment about my discussion of the citation background for U.S. 3,900,554, issued on Aug. 19, 1975, "Method for the reduction of the concentration of NO in combustion effluents using ammonia" by Richard K. Lyon. In July, 1996, we noted that the patent had been cited in the "references cited" field of 92 patents, and in 133 patents overall. By October, 1996, the numbers had grown respectively to 98 and 141, with none of the recent patents by the assignee company. Although BN noted "a company that does not build on the prior art and technology that it has already patented would be one with a totally disorganized R&D effort", such arguably is the case with the '554 patent [see also Physics Today, pp. 59-60 (July 1987); Business Week, pp. 72-75 (April 2, 1990); Carbon 1995, 33, 1007-1010; Nature, 382, 17-18 (July 4, 1996)].
BN made no comments on the letters by Louis Robertson and Steve Mendelsohn in the July, 1996 issue of The Law Works. Furthermore, they have not fully responded to Paul Wille's argument in the April, 1996 issue. The patents that are listed in the "references cited" field of the first page of a patent originate from two sources: the examiner (PTO-892 form) and the applicant (PTO-1449) form. Of the first source, the examiner may cite a patent for what it teaches. There can be many uses made by the examiner of a patent reference. For example, I recently had an examiner cite a patent for its discussion of the prior art, not for the invention of the patent. Of the second source, the applicant is free to list patents for consideration by the patent office; as long as they comply with the rules, they will appear in the "references cited" field of the patent. There is no barrier to self-citation by the applicant. It is not clear that the method of BN distinguishes between references cited on a PTO-892 form and on a PTO-1449 form.
BN promise to review all the studies which have shown the importance of highly cited patents. There are many papers that have questioned the value of patent citation. n12
n12 Edlyn S. Simmons and Nancy Lambert, "Patent Statistics: Comparing Grapes and Watermelons", Proc. Montreux 1991 Int. Chemical Information Conference published in "Recent Advances in Chemical Information", Royal Society of Chemistry, Spec. Publ. 100, 1992.; see also Stuart M. Kaback, Nancy Lambert, and Edlyn S. Simmons, "Patent Citations: Source of Insight or Nothing to Get Excited About?", August 1994 meeting of American Chemical Society [this reference notes: "Patent citations turn out to be quite different in character from the citations we may be familiar with in the citation literature."]; E. Simmons, "Patent family databases 10 years later", Database, Vol. 18(3), Pg. 28, (June, 1995).
**One can find more than one paper criticizing patent citation analysis if one seriously looks.