PTAB

IPR2014-00320

Apple Inc. v. RENSSELAER POLYTECHNIC INSTITUTE

Log In

1. Case Identification

2. Patent Overview

  • Title: Natural Language Interface Using Constrained Intermediate Dictionary of Results
  • Brief Description: The ’798 patent discloses methods and systems for processing natural language queries to search databases. The invention purports to allow a user to input a query in natural language, which is then processed without user augmentation to identify database objects and determine combinations to form a query result.

3. Grounds for Unpatentability

Ground 1: Anticipation by Bouchou - Claims 1-6, 9-12, 14-17, and 20-21 are anticipated under 35 U.S.C. §102 by Bouchou.

  • Prior Art Relied Upon: Bouchou et al., Using Transducers in Natural Language Database Query, 4th International Conference on Applications of Natural Language to Information Systems (“Bouchou”).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Bouchou, which describes a system for natural language database querying, discloses every limitation of the challenged claims. Bouchou’s system uses a "cascade of sequential transducers" and an "electronic dictionary" that Petitioner contended meet the limitations of the claimed "language-based databases" and "metadata database," respectively. The dictionary in Bouchou is built from database structure, contents, and linguistic operators, allegedly teaching a metadata database containing case information, keywords, information models, and database values. Petitioner asserted that Bouchou’s process of searching its dictionary based on the user's input without user supplementation teaches the "without augmentation" limitation. The subsequent processing by transducers to rearrange codes and generate a formal SQL query was argued to teach the claim elements of identifying database objects and determining a plurality of combinations.
    • Key Aspects: Petitioner contended that Bouchou's system, which processes natural language (French) and resolves ambiguities using linguistic operators, directly maps to the methods described in both independent claims 1 and 9 and their respective dependent claims.

Ground 2: Anticipation and Obviousness over Dar - Claims 1-11 and 13-21 are anticipated under 35 U.S.C. §102 by Dar, and Claim 12 is obvious under 35 U.S.C. §103 over Dar.

  • Prior Art Relied Upon: Dar, DTL’s DataSpot: Database Exploration Using Plain Language (“Dar”), and the general knowledge of a person of ordinary skill in the art (POSITA).

  • Core Argument for this Ground:

    • Prior Art Mapping: Petitioner argued that Dar describes a database tool, DataSpot, that anticipates nearly all challenged claims. Dar’s "hyperbase," a graph-based data structure storing linguistic information, user-defined associations, and data objects, was asserted to be a "metadata database" as claimed. Petitioner contended that Dar’s system accepts "free-form" natural language queries and processes them without user supplementation, thus teaching the "without augmentation" limitation. The process in Dar of identifying "hyperbase nodes" based on the query and then searching for connections between them was argued to teach the limitations of identifying database objects and determining combinations. For dependent claims, Petitioner asserted that Dar's use of heuristics to rank results and resolve ambiguity taught the corresponding limitations.
    • Motivation to Combine (for Claim 12): While arguing Dar anticipated most claims, Petitioner contended that even if Dar did not explicitly disclose formulating a query in SQL (as required by claim 12), it would have been obvious for a POSITA to do so. SQL was a well-known, standard query language for relational databases at the time. A POSITA implementing a system like Dar would combine it with a standard tool like SQL to achieve the predictable result of executing a formal database query.
    • Expectation of Success (for Claim 12): A POSITA would have had a high expectation of success, as combining a natural language front-end with a standard SQL back-end was a common and well-understood practice for database interface design.

  • Additional Grounds: Petitioner asserted additional anticipation and obviousness challenges against claims 1-21 based on Warthen (Patent 6,584,464). The arguments relied on similar theories, mapping Warthen's "knowledge base" and "semantic net snapshot" to the claimed "metadata database" and its query processing steps to the claimed method limitations.

4. Key Claim Construction Positions

  • "Without Augmentation": Petitioner argued this term should be construed as "based on the natural language input provided by the user, without user supplementation of the input through the addition of information or structure." Petitioner adopted this construction, which was proposed by the Patent Owner in co-pending litigation, to demonstrate that the prior art meets even the Patent Owner’s own interpretation of the term.
  • "Metadata Database": Petitioner proposed construing this term as a "database including information about structuring, using, or interpreting data." This construction, also adopted from the Patent Owner's litigation positions, was argued to be consistent with the specification and broad enough to read on the "electronic dictionary" of Bouchou, the "hyperbase" of Dar, and the "knowledge base" of Warthen.

5. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1-21 of the ’798 patent as unpatentable.