PTAB
IPR2025-01336
Meta Platforms Inc v. Dialect LLC
Key Events
Petition
Table of Contents
petition
1. Case Identification
- Case #: IPR2025-01336
- Patent #: 9,734,825
- Filed: August 15, 2025
- Petitioner(s): Meta Platforms, Inc.
- Patent Owner(s): Dialect, LLC
- Challenged Claims: 5-8
2. Patent Overview
- Title: Method for Responding to a User Generated Natural Language Speech Utterance
- Brief Description: The ’825 patent describes systems and methods for processing natural language speech by determining a domain for a user's utterance based on its content and context. The system uses autonomous, executable "domain agents" to respond to queries and commands.
3. Grounds for Unpatentability
Ground 1: Claims 5-7 are obvious over Ross in view of Ittycheriah.
- Prior Art Relied Upon: Ross (Patent 7,085,723) and Ittycheriah (Patent 5,937,383).
- Core Argument for this Ground:
- Prior Art Mapping: Petitioner argued that Ross disclosed a multi-context speech application system that met most limitations of independent claim 5, such as determining a domain for a spoken utterance and selecting a domain agent (an "external application" in Ross) to respond. Ross, however, did not detail the underlying "speech engine." Petitioner asserted that Ittycheriah supplied the missing details, teaching a speech engine that recognizes keywords and calculates associated likelihood scores (probabilities) using a cache of recently used words, fulfilling the limitations of claim 5[a] and 5[b]. Ross's "context manager" was alleged to function as the claimed parser, which receives the keyword and probabilities from the Ittycheriah engine and then determines a score for possible contexts based on recency of access, satisfying limitations 5[c] and 5[d].
- Motivation to Combine: A POSITA would combine Ross with Ittycheriah because Ross explicitly taught using commercially available, off-the-shelf speech engines, particularly those from IBM. Ittycheriah was an IBM-developed speech recognition patent that provided a known and compatible method for implementing Ross's speech engine. Furthermore, Ittycheriah’s use of a cache database to speed up recognition would have been a desirable improvement, aligning with the general goal of processing speech commands quickly and efficiently.
- Expectation of Success: A POSITA would have had a high expectation of success because both references were developed by IBM and operated in the same technical field. The combination involved integrating a known type of speech engine (Ittycheriah) into a system designed to use such engines (Ross), a predictable integration of known technologies.
Ground 2: Claims 5-8 are obvious over Ross in view of Ittycheriah and Glover.
- Prior Art Relied Upon: Ross (Patent 7,085,723), Ittycheriah (Patent 5,937,383), and Glover (Application # 2002/1065860).
- Core Argument for this Ground:
- Prior Art Mapping: This ground incorporated the arguments from Ground 1 for claims 5-7. For the additional limitations of dependent claim 8, Petitioner argued that Glover taught a metasearch engine that creates and sends duplicate queries to different information sources asynchronously. Glover's "dispatcher" was alleged to send a user's query to multiple underlying search engines in parallel. This functionality directly mapped to claim 8's requirement that a domain agent be configured to "create duplicate queries and send the duplicate queries in an asynchronous manner to different information sources."
- Motivation to Combine: While Ross’s system involved external applications performing commands, it did not detail how those applications would source information from the internet. A POSITA seeking to implement robust internet search functionality within Ross's framework would combine it with a known metasearch engine like Glover. Glover offered a method to improve search coverage by querying multiple sources, a known benefit that would have been desirable to implement in Ross's system.
- Expectation of Success: Success was predictable, as it involved extending the functionality of Ross's domain agents with a known network interface to connect to Glover's dispatcher. This was a straightforward implementation of internet search capability into a speech-command system.
Ground 3: Claims 5-8 are obvious over Hartono in view of Baker.
Prior Art Relied Upon: Hartono (WO 00/11571) and Baker (Patent 5,680,511).
Core Argument for this Ground:
- Prior Art Mapping: Petitioner contended that Hartono disclosed a modular, agent-based architecture for processing natural language commands. Like Ross, Hartono suggested using any "suitable off the shelf speech recognition system." Petitioner argued that Baker disclosed such a system, which used both acoustic models and contextual "word lists" to generate probability and rank signals for recognized words, thus satisfying limitations 5[a]-5[c]. In this combination, Baker’s speech system acted as the "system agent" providing keywords and probabilities to Hartono’s "natural language syntactic parser" and "semantic interpreter," which determined the appropriate domain and selected a corresponding task agent. Hartono's disclosure that its task agents for different applications (e.g., Netscape and Microsoft Exchange) could execute the same conceptual task was argued to teach the "duplicate queries" of claim 8.
- Motivation to Combine: A POSITA would combine Hartono and Baker because Hartono explicitly contemplated using a commercially available speech recognition system. Baker, developed by the well-known Dragon Systems, represented a suitable and powerful off-the-shelf option. Baker’s ability to recognize words "on the fly" without extensive data training would have been seen as a fast and efficient way to implement the speech recognition module in Hartono's system.
- Expectation of Success: A POSITA would have expected success in combining the references, as it involved integrating a known type of speech recognition engine (Baker) into a system designed for such a component (Hartono). Baker’s system was designed to output recognized words for use by other applications, making it compatible with Hartono’s parsing and interpretation modules.
Additional Grounds: Petitioner asserted an additional obviousness challenge for claim 8 over Hartono, Baker, and Glover, relying on Glover to more explicitly teach the creation and asynchronous sending of duplicate queries to different internet sources.
4. Relief Requested
- Petitioner requests institution of an inter partes review and cancellation of claims 5-8 of Patent 9,734,825 as unpatentable.
Analysis metadata