PTAB

IPR2020-01390

Amazon.com Inc v. VB Assets LLC

Log In
Key Events
Petition
petition Intelligence

1. Case Identification

2. Patent Overview

  • Title: System and Method for Selecting and Presenting Advertisements Based on Natural Language Processing of Voice-Based Input
  • Brief Description: The ’176 patent describes a system for selecting and presenting advertisements in response to voice-based user input. The system uses a speech recognition engine and a conversational language processor to receive a natural language utterance, interpret its meaning to establish a context, and then select and present a relevant advertisement.

3. Grounds for Unpatentability

Ground 1: Obviousness over Kennewick, Yonebayashi, and Jong - Claims 1-3, 6-19, 22-29, 32-45, and 48-52 are obvious over Kennewick in view of Yonebayashi and Jong.

  • Prior Art Relied Upon: Kennewick (Application # 2004/0193420), Yonebayashi (Japanese Application # 2002-297626), and Jong (Patent 6,173,250).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Kennewick and Yonebayashi collectively disclose all elements of the base claims. Kennewick taught a "mobile interactive natural language speech system" that receives voice queries, processes the input to determine context (using user profiles, location, etc.), and provides promotional offers. Yonebayashi similarly taught a speech recognition system that determines a topic from a user's speech and presents a corresponding advertisement. The specific limitation of claim 1 requiring the mapping of phonemes to syllables to recognize words was argued to be taught by Jong, which discloses a speech recognition device that identifies phonemes, extracts corresponding syllables, and checks them against a lexicon database to group them into recognizable words.
    • Motivation to Combine: A POSITA would combine Kennewick and Yonebayashi, as they are in the same field of interactive speech technology and provide complementary disclosures for presenting advertisements based on voice input. Specifically, a POSITA would refine Kennewick's system using Yonebayashi's teachings on selecting ads based on user interests and dialogue history. A POSITA would further be motivated to implement the detailed speech recognition architecture of Jong into the combined Kennewick/Yonebayashi system to enhance how words and phrases are recognized from a user's voice input.
    • Expectation of Success: A POSITA would have a reasonable expectation of success because the references describe similar architectures for speech recognition and advertisement selection, and combining them would involve applying known techniques to improve a known system.

Ground 2: Obviousness over Kennewick, Yonebayashi, Jong, and Colledge - Claims 4, 5, 20, 21, 30, 31, 46, and 47 are obvious over the combination of Kennewick, Yonebayashi, and Jong in further view of Colledge.

  • Prior Art Relied Upon: Kennewick (Application # 2004/0193420), Yonebayashi (Japanese Application # 2002-297626), Jong (Patent 6,173,250), and Colledge (Patent 7,774,333).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground builds upon the combination in Ground 1. Petitioner asserted that the additional limitations of the challenged claims, such as in claim 4 ("builds statistical profiles for selecting subsequent advertisements"), were taught by Colledge. Colledge disclosed a method of providing advertisements by tracking user interaction (viewing habits, purchasing habits, demographics) and using a machine learning classifier (e.g., a Naïve Bayes classifier) to build statistical profiles. These profiles, which represent computed probabilities of a match between user characteristics and ad characteristics, were then used to select subsequent advertisements likely to interest the user.
    • Motivation to Combine: A POSITA would be motivated to modify the system from Ground 1 with the teachings of Colledge to increase the effectiveness of advertisement selection. By incorporating Colledge’s use of a machine learning classifier to build statistical profiles based on tracked user data, the combined system would be more likely to select and present advertisements that match user preferences and characteristics, thereby enhancing the overall performance of the system.
    • Expectation of Success: A POSITA would have expected success in this combination, as it involved integrating a known machine learning technique (from Colledge) for improving advertisement targeting into an established voice-based advertisement system (from Kennewick/Yonebayashi/Jong).

4. Arguments Regarding Discretionary Denial

  • Petitioner argued that the Board should not exercise its discretion to deny institution under §314(a) or §325(d).
  • Regarding co-pending litigation, Petitioner noted that it challenged all claims (1-52) in the IPR, whereas the district court complaint only alleged infringement of claim 27. Denying institution would therefore deprive Petitioner of an alternative forum to challenge the validity of the remaining 51 claims.
  • Regarding §325(d), Petitioner argued that while Kennewick was cited during prosecution, it was part of an Information Disclosure Statement (IDS) that listed over 120 references and was not substantively considered by the examiner. Petitioner asserted that the Board has consistently declined to deny institution in such circumstances and that the Becton factors do not favor denial.

5. Relief Requested

  • Petitioner requests institution of inter partes review and cancellation of claims 1-52 of the ’176 patent as unpatentable.