Google LLC v. Valtrus Innovations Limited

1. Case Identification

• Case #: IPR2022-01545
• Patent #: 6,728,704
• Filed: September 15, 2022
• Petitioner(s): Google LLC
• Patent Owner(s): Valtrus Innovations Limited
• Challenged Claims: 1-23

2. Patent Overview

• Title: Method for Merging Result Lists from Multiple Search Engines
• Brief Description: The ’704 patent discloses a method for arranging web search results from multiple search engines into a single merged list. The system assigns a "representative value" to each search engine's result list, and then uses this value to iteratively or probabilistically select entries from the various lists to create a final, unified ranking for the user.

3. Grounds for Unpatentability

Ground 1: Obviousness over Bushee and Voorhees - Claims 1-3, 5-9, 12-14, 17-20, and 23 are obvious over Bushee in view of Voorhees.

• Prior Art Relied Upon: Bushee (Patent 6,711,569) and Voorhees (Patent 5,864,846).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Bushee taught a multi-search-engine system that obtains a query, receives results, selects a subset, scores them, and calculates an average score (a representative value) for each search engine. However, Bushee's method for presenting results was rudimentary, simply placing the results from each engine in blocks ranked by the engine's average score. Petitioner argued that Voorhees taught a superior method for "fusing" results from multiple search engines into a single interleaved list using a representative value (the number of pages to retrieve) that varies as results are selected and added to the final list.
  • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine the references to improve the rudimentary result presentation of Bushee. Implementing Voorhees's more sophisticated "fusion" method within Bushee's system would achieve the known benefit of presenting a single, properly ranked list to the user, making the use of multiple search engines transparent and more effective.
  • Expectation of Success: A POSITA would have had a reasonable expectation of success as the combination involved applying a known, improved result-merging technique (Voorhees) to a known multi-engine search system (Bushee) to achieve a predictable improvement.

Ground 2: Obviousness over Bushee, Voorhees, and Koppel - Claims 3-4, 9-10, 14-15, and 20-21 are obvious over Bushee in view of Voorhees and Koppel.

• Prior Art Relied Upon: Bushee (Patent 6,711,569), Voorhees (Patent 5,864,846), and Koppel (Patent 7,257,766).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground built upon Ground 1 and introduced Koppel to address dependent claims requiring specific methods for selecting a subset of results (e.g., selecting entries that are "uniformly spaced" or "at random"). While Bushee taught selecting a subset (the "first N results"), Koppel explicitly disclosed various known methods for selecting a smaller number of search results from a larger set, including selecting "evenly...spaced sites" (mapping to claim 3) and "randomly selecting sites" (mapping to claim 4).
  • Motivation to Combine: A POSITA would have been motivated to use the methods in Koppel as simple, well-known, and alternative design choices for implementing the "selecting a subset" step already present in Bushee. The ’704 patent placed no particular emphasis on the specific selection method, suggesting these were simple substitutions.
  • Expectation of Success: Implementing the selection methods of Koppel into the Bushee/Voorhees system would have been a simple substitution of one known technique for another, leading to predictable results.

Ground 3: Obviousness over Voorhees, Bushee, and Tso - Claims 1-3, 5-9, 12-14, 17-20, and 23 are obvious over Voorhees in view of Bushee and Tso.

• Prior Art Relied Upon: Voorhees (Patent 5,864,846), Bushee (Patent 6,711,569), and Tso (Patent 6,385,602).

• Core Argument for this Ground:

  • Prior Art Mapping: This ground positioned Voorhees as the primary reference, arguing it would have been obvious to modify its method for calculating search-engine weights. Petitioner contended that a POSITA would have used an average score of the search results, as taught by both Bushee and Tso, as the weight in Voorhees's fusion algorithm. Bushee taught calculating an average score for each database to evaluate its relevance. Tso redundantly taught calculating a "category relevance rating" by assigning the "average score of all items in the category."
  • Motivation to Combine: A POSITA would have been motivated to use the average score method of Bushee/Tso because it was simpler and more adaptive than Voorhees's method, which relied on pre-training queries. Using an average score is dependent only on the results of the current query, avoiding the need for continual re-processing of training data and better representing the immediate relevance of a search engine's results.
  • Expectation of Success: Substituting a well-known weighting metric (average score) for another in Voorhees's known system was a predictable modification that would have offered several advantages, such as simplicity and improved real-time relevance.

• Additional Grounds: Petitioner asserted an additional obviousness challenge (Ground 4) against claims 3-4, 9-10, 14-15, and 20-21 based on the combination of Voorhees, Bushee, Tso, and Koppel, applying the logic from Ground 2 to the combination established in Ground 3.

4. Arguments Regarding Discretionary Denial

• §325(d) - Same or Substantially Same Art/Arguments: Petitioner argued against denial under §325(d), stating that the principal references (Bushee, Voorhees, Tso, and Koppel) were not of record during the prosecution of the ’704 patent. While two publications by the inventor of Voorhees were of record, Petitioner asserted they were not the same as the asserted Voorhees patent and lacked the clear technical disclosure necessary to render the claims obvious.
• §314(a) - Fintiv Factors: Petitioner argued that the Fintiv factors weighed in favor of institution. The petition was filed early in the parallel district court litigation, with discovery just beginning and no trial date set. Given the court's median time-to-trial, the Board would likely reach a Final Written Decision before or around the same time as a trial on validity. Finally, Petitioner contended that the merits of the petition presented compelling evidence of unpatentability, which should favor institution.

5. Relief Requested

• Petitioner requests institution of inter partes review and cancellation of claims 1-23 of Patent 6,728,704 as unpatentable.