Microsoft Corp v. IPA Technologies Inc

1. Case Identification

• Case #: IPR2018-00734
• Patent #: 6,523,061
• Filed: March 6, 2018
• Petitioner(s): Microsoft Corporation
• Patent Owner(s): IPA Technologies, Inc.
• Challenged Claims: 1-18

2. Patent Overview

• Title: System, Method, and Article of Manufacture for Agent-Based Navigation in a Speech-Based Data Navigation System
• Brief Description: The ’061 patent describes a system and method for navigating electronic data sources using spoken natural language. The system employs a platform of distributed software agents that work collaboratively to receive a spoken request, interpret it, construct a navigation query, and retrieve the desired information for the user.

3. Grounds for Unpatentability

Ground 1: Claims 1-18 are anticipated under 35 U.S.C. §102 by Moran.

• Prior Art Relied Upon: Moran (a 1997 publication titled Multimodal User Interfaces in the Open Agent Architecture).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Moran, a publication co-authored by two of the ’061 patent’s inventors, discloses every element of the challenged claims. Moran describes a multimodal, agent-based user interface built on the same Open Agent Architecture (OAA) platform disclosed in the ’061 patent. Moran’s system receives spoken commands (“receiving a spoken request”), uses a User Interface Agent to manage inputs, and delegates tasks to specialized agents, including a speech recognition agent and a natural language understanding agent, to determine the user's intent (“rendering an interpretation”). This interpretation is converted into a high-level logical language (ICL), which functions as a “navigation query.” A "Facilitator" agent then routes this query to the appropriate application agents (e.g., database agents) that access electronic data sources to retrieve the requested information.
  • Key Aspects: The core of this argument is that the patent claims technology its own inventors had already published more than a year before the patent's filing date, constituting a statutory bar to patentability under §102(b).

Ground 2: Claims 1-18 are obvious under 35 U.S.C. §103 over Moran, with or without Kistler.

• Prior Art Relied Upon: Moran (a 1997 publication), Kistler (a 1998 publication titled WebL — a programming language for the Web).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground was presented as an alternative to anticipation. For claim limitations not explicitly found in Moran, Petitioner argued they would have been obvious. Specifically for claims 6, 12, and 18, which require an agent that "scrapes the web page," Petitioner contended that while Moran discloses a "Web retrieval agent" that scans online databases, Kistler explicitly teaches the concept of web scraping. Kistler describes a programming language (WebL) designed to extract data from web pages, which it views as a distributed database.
  • Motivation to Combine: A POSITA seeking to implement Moran's concept of a "Web retrieval agent" to access unstructured online sources like webpages would naturally look to known techniques for data extraction. Kistler provides precisely such a technique. Moran itself encourages a "mix-and-match" approach to incorporating pre-existing components into its OAA framework, motivating the integration of a web-scraping agent based on teachings like Kistler's to enhance the system's data retrieval capabilities.
  • Expectation of Success: A POSITA would have a reasonable expectation of success in combining these teachings, as it would involve supplementing Moran's agent architecture with a known method for data retrieval from a specific type of data source (webpages), a task well within the ordinary skill in the art.

Ground 3: Claims 1-18 are obvious under 35 U.S.C. §103 over Moran in view of Burns, with or without Kistler.

• Prior Art Relied Upon: Moran (a 1997 publication), Burns (Patent 5,454,106), and optionally Kistler (a 1998 publication).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground addresses any alleged deficiency in Moran regarding the specific step of “constructing a navigation query” from a parsed interpretation of a user's request. Petitioner argued that Burns explicitly discloses this element. Burns describes a database retrieval system that uses natural language input, parses it, and explicitly discloses converting the parsed results into a structured database query to retrieve information. This combination fully teaches the claimed method. The addition of Kistler serves the same purpose as in Ground 2 for the web-scraping limitations.
  • Motivation to Combine: A POSITA implementing Moran’s high-level agent architecture would have been motivated to incorporate the well-established database query construction and retrieval methods taught by Burns. Moran's system requires a mechanism to translate user intent into a machine-executable query, and Burns provides a detailed, conventional solution for doing so. Moran’s encouragement of incorporating pre-existing components would lead a POSITA to Burns to implement the practical details of database interaction within the broader OAA framework.
  • Expectation of Success: A POSITA would reasonably expect success in integrating Burns's query generation logic into Moran's agent-based system. This would involve creating an agent that performs the functions described in Burns, a straightforward implementation task that leverages Moran’s flexible, modular architecture.

4. Key Claim Construction Positions

• "navigation query": Petitioner referenced Patent Owner's proposed construction from a related district court case: “an electronic query, form, series of menu selections, or the like; being structured appropriately so as to navigate a particular data source of interest in search of desired information.” Petitioner argued Moran’s use of a high-level logical language (ICL) to represent a user's command for automated agents met this construction.
• "rendering an interpretation of the spoken request": Referencing the Patent Owner's proposal, this term was construed as “determining a meaning of the spoken request using a computing device, such as that provided by extracting speech data from acoustic voice signals or data and linguistically parsing the speech data.” Petitioner argued Moran’s use of a speech recognition agent followed by a natural language understanding agent directly satisfied this construction.

5. Relief Requested

• Petitioner requested the institution of an inter partes review and the cancellation of claims 1-18 of the ’061 patent as unpatentable.