PTAB

IPR2025-00871

Samsung Electronics Co Ltd v. VB Assets LLC

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Facilitating Natural Language System Responses Utilizing Accumulated Short-Term and Long-Term Knowledge
  • Brief Description: The ’341 patent discloses a cooperative conversational voice user interface. The system processes user utterances by utilizing both short-term knowledge (from the current conversation) and long-term knowledge (from past conversations) to determine context, interpret intent, and generate a response.

3. Grounds for Unpatentability

Ground 1: Obviousness over SmartKom and Kobsa - Claims 1-18 are obvious over SmartKom in view of Kobsa.

  • Prior Art Relied Upon: SmartKom (a 2006 book titled "SmartKom: Foundations of Multimodal Dialogue Systems") and Kobsa (a 1989 book titled "User Models in Dialog Systems").
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that SmartKom, a multimodal dialogue system, discloses the core elements of the challenged claims, including a system that processes user utterances. SmartKom’s "dialogue model" and "discourse state," which track the history of the current conversation, were alleged to teach the claimed "short-term knowledge." Petitioner contended that SmartKom’s use of a "user model" to store user preferences and properties teaches the concept of "long-term knowledge." However, SmartKom provides limited detail on implementing these user models. Kobsa, a comprehensive survey on user modeling, was argued to supply these missing details. Specifically, Kobsa teaches building persistent, long-term user models that store user preferences and are updated after a dialogue session ends. Petitioner asserted that the combination teaches expiring short-term knowledge (the discourse state of a completed dialogue) and accumulating it into long-term knowledge (the persistent user model), as required by the claims.
    • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine these references because SmartKom explicitly suggests using user models and profiles to provide tailored responses but lacks implementation details. A POSITA would be motivated to consult a foundational text like Kobsa to implement this known technique effectively. The goal would be to improve the dialogue system's user experience by making it more intelligent and cooperative, a benefit stressed by Kobsa. The fact that an editor of SmartKom also co-edited Kobsa would have further directed a POSITA to the reference.
    • Expectation of Success: A POSITA would have a reasonable expectation of success because the combination involves integrating known user modeling software constructs (from Kobsa) into a system with a standard software architecture (SmartKom).

Ground 2: Obviousness over Barbara and Ross - Claims 1-18 are obvious over Barbara in view of Ross.

  • Prior Art Relied Upon: Barbara (Application # 2004/0101198) and Ross (Application # 2002/0173960).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Barbara discloses a voice interface system that interprets spoken utterances by capturing contextual information. This contextual information includes both recent dialogue history (short-term knowledge) and accumulated user profile information like preferences from past interactions (long-term knowledge). However, Barbara was argued to lack detail regarding its server architecture and data management. Ross allegedly supplies these details by teaching a "conversation manager" that uses a "conversational record" to store the dialogue history for a current session. This record is eventually purged when no longer relevant to active goals, which Petitioner mapped to the claim limitation of "expiring" short-term knowledge. Information from these expired records could then be stored in a permanent user information database, as taught by Barbara, satisfying the claim requirement that expired short-term knowledge becomes part of the accumulated long-term knowledge.
    • Motivation to Combine: A POSITA would be motivated to combine the references to improve Barbara's system. Barbara describes a functional voice interface but is vague on implementation. A POSITA seeking to build such a system would look to references like Ross for its robust disclosure of a conversation manager architecture, response generation capabilities, and efficient data management. Specifically, implementing Ross’s technique of purging conversational records would improve Barbara's system by increasing storage efficiency, a known benefit for such systems.
    • Expectation of Success: A POSITA would expect success because the combination involves applying a known data management technique (Ross’s expiring conversational records) to a known type of device (Barbara’s voice interface system) to achieve a predictable improvement in efficiency.

4. Arguments Regarding Discretionary Denial

  • To simplify the analysis under Fintiv, Petitioner stipulated that if the IPR is instituted, it will not pursue in the related district court proceeding any ground that it raised or reasonably could have raised against the challenged claims in the IPR.

5. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1-18 of the ’341 patent as unpatentable.