Rensselaer Polytechnic Institute v. Amazon.com, Inc.

CourtDistrict Court, N.D. New York
DecidedMarch 18, 2024
Docket1:18-cv-00549
StatusUnknown

This text of Rensselaer Polytechnic Institute v. Amazon.com, Inc. (Rensselaer Polytechnic Institute v. Amazon.com, Inc.) is published on Counsel Stack Legal Research, covering District Court, N.D. New York primary law. Counsel Stack provides free access to over 12 million legal documents including statutes, case law, regulations, and constitutions.

Bluebook
Rensselaer Polytechnic Institute v. Amazon.com, Inc., (N.D.N.Y. 2024).

Opinion

UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF NEW YORK

RENSSELAER POLYTECHNIC INSTITUTE and CF DYNAMIC ADVANCES LLC, 1:18-cv-549 (BKS/TWD) Plaintiffs,

v.

AMAZON.COM, INC.,

Defendant.

Appearances: For Plaintiffs: Ronald J. Schutz Cyrus A. Morton Benjamen C. Linden Navin Ramalingam Robins Kaplan LLP 800 LaSalle Avenue, Suite 2800 Minneapolis, MN 55402

Annie Huang Miles Finn Robins Kaplan LLP 900 Third Avenue, Suite 1900 New York, NY 10022

Li Zhu Robins Kaplan LLP 555 Twin Dolphin Drive, Suite 310 Redwood City, CA 94065

Lucas M. Walker MoloLamken LLP 600 New Hampshire Avenue NW, Suite 500 Washington, DC 20037 For Defendant: John G. Powers James P. Youngs Mary L. D’Agostino Hancock Estabrook, LLP AXA Tower I, Suite 1800 100 Madison Street Syracuse, NY 13202

Joseph R. Re Jeremy A. Anapol Knobbe, Martens, Olson & Bear, LLP 2040 Main Street, 14th Floor Irvine, CA 92614

Colin B. Heideman Nathan D. Reeves Knobbe, Martens, Olson & Bear, LLP 925 Fourth Avenue, Suite 2500 Seattle, WA 98104

Yanna S. Bouris Knobbe Martens Olson & Bear, LLP 1925 Century Park East, Suite 400 Los Angeles, CA 90067

Jon W. Gurka Crowell & Moring LLP 3 Park Plaza, 20th Floor Irvine, CA 92614 Hon. Brenda K. Sannes, Chief United States District Judge: MEMORANDUM-DECISION AND ORDER I. INTRODUCTION Plaintiffs Rensselaer Polytechnic Institute (“RPI”) and CF Dynamic Advances LLC (“CF Dynamic”) bring this patent infringement action against Defendant Amazon.com, Inc. (“Amazon”). (Dkt. No. 1). Presently before the Court are the parties’ motions for summary judgment. (Dkt. Nos. 384, 390). The parties have filed responsive briefing with respect to both motions, (Dkt. Nos. 400, 407, 413, 414), and the Court heard oral argument on March 14, 2024, (Text Minute Entry, March 14, 2024). Also before the Court are the parties’ related motions to strike or otherwise exclude expert testimony and statements of material fact. (Dkt. Nos. 383, 386, 388, 390, 411). These motions are fully briefed as well. (Dkt. Nos. 400, 401, 403, 405, 413, 416, 418, 420). For the following reasons, Plaintiffs’ motion for summary judgment is denied, Defendant’s motion for summary judgment is granted, and the parties’ related motions are denied as moot.

II. BACKGROUND1 A. The Patent The asserted patent in this infringement action, United States Patent No. 7,177,798 (the “’798 Patent”), is titled “Natural Language Interface Using Constrained Intermediate Dictionary of Results.” ’798 Patent, col. 1 ll. 1–3. The ’798 Patent issued on February 13, 2007, and lists Cheng Hsu and Veera Boonjing as the inventors and RPI as the assignee. (See Dkt. No. 89-2 (2001 assignment of all rights from the inventors to RPI)). CF Dynamic holds exclusionary rights to the ’798 Patent. See Rensselaer Polytechnic Inst. v. Amazon.com, Inc., No. 18-cv-549, 2023 WL 6037877, at *1, 2023 U.S. Dist. LEXIS 163947, at *2 (N.D.N.Y. Sept. 15, 2023). The ’798 Patent relates to the field of natural language processing (“NLP”). ’798 Patent, col. 1 ll. 22–24. Research in this field aims to enable humans to interact with machines or

computers using human natural language rather than computer languages or specialized commands. Id. at col. 4 ll. 49–65. But the ambiguity inherent in natural language presents a challenge. Id. at col. 5. ll. 19–42. To interpret and process a natural language query, a natural language interface (“NLI”) must understand the natural articulation of the query or enumerate and evaluate all possible interpretations of it. Id. at col. 5 ll. 24–26. And because in either case

1 The facts are construed in the light most favorable to the non-moving party. See Gilles v. Repicky, 511 F.3d 239, 243 (2d Cir. 2007). That is, the facts are construed in the light most favorable to Defendant when the Court considers Plaintiffs’ motion, and the facts are construed in the light most favorable to Plaintiffs when the Court considers Defendant’s motion. See Lumbermens Mut. Cas. Co. v. RGIS Inventory Specialists, LLC, 628 F.3d 46, 51 (2d Cir. 2010). experience can be used to improve interpretation, an effective NLI must learn from past performance. Id. at col. 5 ll. 27–30. The ’798 Patent details five approaches to NLIs known at the time of the invention. See id. at col. 5 l. 43 to col. 6 l. 34. Four of these approaches—the template-based approach, syntax-

based approach, semantic-grammar-based approach, and intermediate-representation-language- based approach—restrict the syntax with which a user can pose a natural language query. Id. at col. 5 l. 60 to col. 6 l. 2. The user must articulate a query using only sentence structures and words known by the NLI. Id. at col. 4 ll. 58–61, col. 5 l. 60 to col. 6 l. 2. Properly articulated queries are translated into standard database query language and processed. Id. at col. 5 ll. 43–53. But improperly articulated queries are often interpreted and processed inaccurately, if at all. Id. at col. 5 ll. 60–66. The fifth approach—the semantic-model-based approach—does not so depend on the rigid construction of a query. Id. at col. 6 ll. 3–5. Instead, keywords in a query are identified and associated with database objects in a semantic model, with the semantic model acting as a

roadmap to generate possible interpretations of the query. Id. at col. 6 ll. 8–10. However, “[b]ecause a database object can be only a grossly simplistic element of the natural vocabulary, keywords must shoulder the burden of representing naturalness.” Id. at col. 6 ll. 28–31. Therefore, this approach requires a semantic model to predefine an impractically large number of keywords to span the range of possible usages.2 Id. at col. 6 ll. 28–34. The ’798 Patent claims to offer an improved approach, paving the way for NLIs capable of generating all possible interpretations of a natural language query in a form suitable for

2 Defendant’s expert disputes the patent’s assessment that the semantic-model-based approach leads to the exponential growth of keywords. (Dkt. No. 384-3, at 24). That dispute is immaterial to this decision. evaluation while reducing the complexity and growth of keywords. Id. at col. 6 ll. 45–62. The ’798 Patent includes two independent claims—Claims 1 and 9—and nineteen dependent claims. Id. at col. 36 l. 38 to col. 38 l. 37. The two independent claims are similar and each describe “[a] method for processing a natural language input,” though Claim 1 specifies that the input is

provided by a user and Claim 9 specifies that the method is computer-implemented. Id. at col. 36 ll. 38–39, col. 37 ll. 15–16. Claim 1 involves performing—based only on the input and without augmentation—a search of one or more language-based databases, including at least one metadata database, and providing a result of the search to the user. Id. at col. 36 ll. 41–50. Claims 1 and 9 require a metadata database comprised of four types of information: “case information,” “keywords,” “information models,” and “database values.” Id. at col. 36 ll. 43–48, col. 37 ll. 22– 27. Claims 1 and 9 involve identifying a finite number of database objects and determining a plurality of combinations (Claim 1) or permutations (Claim 9) of the database objects. Id. at col. 36 ll. 51–54, col. 37 ll. 18–27. And Claim 9 further involves interpreting at least one permutation to determine a result of the input. Id. at col. 37 ll. 28–29.

Claim 1 recites: 1.

Free access — add to your briefcase to read the full text and ask questions with AI

Related

Anderson v. Liberty Lobby, Inc.
477 U.S. 242 (Supreme Court, 1986)
Jeffreys v. City of New York
426 F.3d 549 (Second Circuit, 2005)
Gilles v. Repicky
511 F.3d 239 (Second Circuit, 2007)
Wright v. Goord
554 F.3d 255 (Second Circuit, 2009)
Ultramercial, Inc. v. Hulu, LLC
772 F.3d 709 (Federal Circuit, 2014)
Ddr Holdings, LLC v. hotels.com, L.P.
773 F.3d 1245 (Federal Circuit, 2014)
Intellectual Ventures I LLC v. Capital One Bank (USA)
792 F.3d 1363 (Federal Circuit, 2015)
Enfish, LLC v. Microsoft Corporation
822 F.3d 1327 (Federal Circuit, 2016)
Electric Power Group, LLC v. Alstom S.A.
830 F.3d 1350 (Federal Circuit, 2016)
McRO, Inc. v. Bandai Namco Games America Inc.
837 F.3d 1299 (Federal Circuit, 2016)
Intellectual Ventures I LLC v. Symantec Corp.
838 F.3d 1307 (Federal Circuit, 2016)
Amdocs (Israel) Limited v. Openet Telecom, Inc.
841 F.3d 1288 (Federal Circuit, 2016)

Cite This Page — Counsel Stack

Bluebook (online)
Rensselaer Polytechnic Institute v. Amazon.com, Inc., Counsel Stack Legal Research, https://law.counselstack.com/opinion/rensselaer-polytechnic-institute-v-amazoncom-inc-nynd-2024.