PTAB

IPR2019-00471

Unified PaTents LLC v. Sisvel Societa ItaLiana per Lo Sviluppo Dell EletTronica Spa

Log In

1. Case Identification

2. Patent Overview

  • Title: Media Browsing Systems and Methods
  • Brief Description: The ’680 patent relates to media browsing systems that generate a user profile based on a user's interests and suggest content based on that profile and the profiles of similar users. The system purports to provide a "virtual unified space" for browsing various media types using implicit, explicit, and collaborative filtering techniques.

3. Grounds for Unpatentability

Ground 1: Claims 1, 4, 8, 10-13, and 16-18 are obvious over Easty in view of Hendricks.

  • Prior Art Relied Upon: Easty (Patent 6,189,008) and Hendricks (Patent 5,798,785).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Easty taught the foundational system of the challenged claims, including an integrated digital content delivery system that creates and continuously updates a user profile. This profile is based on implicit user activities (e.g., content requests, viewing time), explicit information (e.g., demographics), and collaborative data from similar users. Easty further disclosed presenting various media collections, including from broadcast television, in a "virtual unified space." Petitioner asserted that Hendricks supplied the more detailed interface features, teaching a menu-driven graphical interface for browsing categorized content and a user-controlled search engine. Critically, Petitioner contended that Hendricks' method of assigning "weights" to content categories to guide recommendations fulfilled the ’680 patent’s key limitation of updating a user profile where the update is "reflected in a ratio."
    • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine Hendricks' user-friendly menu interface and search functionalities into Easty's foundational system to improve navigation and overall usability, which are known goals in the art. As Easty suggests weighing user preferences, a POSITA would have been motivated to implement a known weighting system like that disclosed in Hendricks. Both systems target similar goals and hardware (e.g., set-top boxes), making the combination a predictable design choice.
    • Expectation of Success: Petitioner asserted a high expectation of success because both references operate in the same technical field of content recommendation. Integrating Hendricks' established interface and weighting techniques into Easty's system would involve routine software modifications yielding the predictable result of an enhanced user experience.

Ground 2: Claims 1, 4, 8, 10-13, and 16-18 are obvious over Easty in view of Hendricks and in further view of Kittaka.

  • Prior Art Relied Upon: Easty (Patent 6,189,008), Hendricks (Patent 5,798,785), and Kittaka (Patent 5,999,975).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground reinforced Ground 1, introducing Kittaka as an alternative or supplementary teaching for the crucial "ratio" limitation, which was the primary basis for the ’680 patent's allowance. Petitioner argued that even if the "weights" in Hendricks were not considered a ratio, Kittaka explicitly disclosed such a mechanism. Kittaka teaches a system that dynamically updates a user's interest profile (represented as a user vector U) based on selected content (an information vector G). The update is performed by calculating a new vector based on an interior division point that divides the line segment between vectors U and G at an express "ratio of C1:1-C1," where C1 is an attenuation coefficient.
    • Motivation to Combine: Petitioner argued that Hendricks’ own disclosure—stating that "more sophisticated methods of weighing the importance of the data may be used"—would have expressly motivated a POSITA to seek out advanced techniques like those in Kittaka. Kittaka's vector-based mathematical approach, which also accounts for the decay of user interest over time, represents such a sophisticated method that would improve the reliability of profile updates in the combined Easty/Hendricks system.
    • Expectation of Success: A POSITA would have had a reasonable expectation of success in this combination. Integrating Kittaka's specific computational model for updating user profiles would be a predictable software enhancement to the known concept of using weighted preferences, leading to more dynamic and accurate content recommendations.

4. Key Claim Construction Positions

  • The petition asserted that the following claim terms, central to the obviousness arguments, should be construed based on the patent’s specification:
    • "implicit filtering": Proposed construction covers providing recommendations based on a user's past, unstated interactions, such as content viewing history, searches performed, and websites visited.
    • "explicit filtering": Proposed construction covers providing recommendations based on a user’s direct input, such as answers to demographic questions or ratings provided for content.

5. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1, 4, 8, 10-13, and 16-18 of the ’680 patent as unpatentable.