DCT

3:23-cv-03955

Ask Sydney LLC v. Google LLC

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 6:23-cv-00111, W.D. Tex., 02/13/2023
  • Venue Allegations: Plaintiff alleges venue is proper in the Western District of Texas because Defendant has regular and established places of business in the District, including an office in Austin, and has committed acts of infringement in the District.
  • Core Dispute: Plaintiff alleges that Defendant’s YouTube and Google Ads platforms infringe patents related to systems and methods for interactively guiding a user to a current interest through a sequence of tagged images.
  • Technical Context: The technology relates to recommendation algorithms that use iterative user feedback to refine search results, a core function in modern e-commerce, content delivery, and online advertising platforms.
  • Key Procedural History: The complaint does not mention any prior litigation, Inter Partes Review (IPR) proceedings, or licensing history related to the patents-in-suit. The '705 Patent is a continuation of the family that produced the '786 Patent, suggesting a related technical focus and a shared prosecution history.

Case Timeline

Date Event
2014-08-15 Earliest Priority Date ('786 Patent)
2014-08-15 Earliest Priority Date ('705 Patent)
2016-04-26 Issue Date ('786 Patent)
2019-11-12 Issue Date ('705 Patent)
2023-02-13 Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 9,323,786 - “System and computer method for visually guiding a user to a current interest” (Issued: April 26, 2016)

The Invention Explained

  • Problem Addressed: The patent describes a deficiency in recommendation systems that rely on a user's broad, long-term history. Such systems are often unable to identify a user's granular, specific, and contemporaneous interest (e.g., a particular food "craving") at a specific moment in time ('705 Patent, col. 1:31-44).
  • The Patented Solution: The invention proposes a guided, iterative method to solve this. A system presents a user with an image representing a physical object (e.g., a food dish), where the image is associated with a set of descriptive "tags." The user provides a preference input (e.g., like/dislike). The system then processes this input and the image's tags to determine a next set of tags, which in turn is used to select the next image to present. This interactive loop continues, theoretically narrowing in on the user's specific "current interest" ('786 Patent, Abstract; col. 2:1-22).
  • Technical Importance: This approach represents a shift from static, history-based recommendations to a dynamic, interactive discovery process designed to resolve a user's immediate, often unarticulated, preferences.

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶18).
  • Essential Elements of Claim 1 (Method):
    • Determining a plurality of tags specific to the user.
    • Transmitting one electronic image to the user's device, with the image being associated with a set of tags.
    • Receiving a preference input from the user for the physical object represented by the image.
    • Processing the tags based on the preference to determine a next set of tags.
    • In response to a negative preference, removing tags from the set of tags available for processing.
    • In response to a positive preference, determining at least one additional tag to add to the current set of tags by determining weightings of other tags.
    • Determining and transmitting a next electronic image based on the next set of tags.
    • Repeating the process to generate a sequence of images.
  • The complaint reserves the right to assert additional claims (Compl. ¶18).

U.S. Patent No. 10,474,705 - “Iterative image search algorithm informed by continuous human-machine input feedback” (Issued: November 12, 2019)

The Invention Explained

  • Problem Addressed: As with the parent '786 Patent, the technology addresses the failure of conventional recommendation systems to predict a user's specific, "in the moment" interest, which may be more granular than their long-term historical preferences ('705 Patent, col. 1:28-56).
  • The Patented Solution: The patent details a computer-implemented image search method that is explicitly "informed by continuous human-machine input feedback." The core loop, illustrated in Figure 2A, involves presenting an image, receiving a user preference input, processing associated tags based on that preference to determine a "next set of tags," and then determining a "next image" based on that new tag set ('705 Patent, Abstract; Fig. 2A). The process is designed to iteratively resolve a user's subjective need ('705 Patent, col. 27:8-12).
  • Technical Importance: This patent formalizes the interactive search concept as an algorithm driven by an explicit human-machine feedback loop, focusing on the adjustment of tag weights based on user input.

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶25).
  • Essential Elements of Claim 1 (Method):
    • Determining a plurality of digital images to present, each associated with a plurality of tags.
    • Receiving an "unfavorable indication of a disinclination" for an image.
    • Analyzing tags to determine a next set of tags.
    • Adjusting weights of tags based on an association relative to tags within the same category.
    • Transitioning the display to a subsequent digital image.
    • Receiving a further input (favorable or unfavorable) for the subsequent image.
  • The complaint reserves the right to assert additional claims (Compl. ¶25).

III. The Accused Instrumentality

Product Identification

The complaint identifies "Google's YouTube, a video sharing and social media platform" and "Google's advertising service Google Ads" as the Accused Instrumentalities (Compl. ¶16).

Functionality and Market Context

  • The complaint identifies YouTube as a video sharing and social media platform and Google Ads as an advertising service (Compl. ¶16).
  • The complaint does not provide specific details on the technical operation of the recommendation or advertising algorithms within these platforms. It broadly alleges that these products are developed, designed, manufactured, and sold by Google and are available to users and businesses throughout the United States (Compl. ¶¶3, 21, 28).

IV. Analysis of Infringement Allegations

The complaint alleges that Google's Accused Products directly infringe at least claim 1 of the '786 Patent and at least claim 1 of the '705 Patent (Compl. ¶¶18, 25). The complaint states that infringement claim charts are attached as Exhibits C and D, respectively (Compl. ¶¶23, 30). However, these exhibits were not filed with the public complaint. The complaint document itself provides no narrative infringement theory or element-by-element mapping.

No probative visual evidence provided in complaint.

The analysis of potential points of contention is therefore based on a juxtaposition of the patent claims against the generally understood functionality of the accused platforms.

Identified Points of Contention ('786 and '705 Patents)

  • Scope Questions:
    • A central question for the court may be whether the term "image," as described and claimed in the patents with a focus on static photographs of physical objects like food dishes ('705 Patent, col. 8:52-56), can be construed to read on video thumbnails (in YouTube) or banner advertisements (in Google Ads).
    • The patents describe a "session of directing the user to the current interest" ('786 Patent, cl. 1) in the context of a dedicated, start-to-finish search. It raises the question of whether the continuous, passive content recommendation on a platform like YouTube constitutes such a "session."
  • Technical Questions:
    • The claims require processing "tags" associated with images to determine a "next" image. A key factual question will be what evidence demonstrates that YouTube or Google Ads uses a comparable "tag"-based iterative refinement system, as opposed to more complex, multi-factorial machine learning models (e.g., collaborative filtering, deep neural networks) for content recommendation and ad selection.
    • Claim 1 of the '786 Patent requires "removing the tags" or "determining at least one additional tag" based on user input. It is an open question what evidence Plaintiff will present to show that Google's algorithms perform these specific positive and negative feedback operations, rather than simply up-weighting or down-weighting a vast array of signals.

V. Key Claim Terms for Construction

  • The Term: "tags" ('786 Patent, cl. 1; '705 Patent, cl. 1)

    • Context and Importance: The definition of "tags" is fundamental to the infringement case. The Plaintiff's infringement theory depends on mapping this term to the data and metadata Google uses to categorize and recommend content. Practitioners may focus on this term because its scope will determine whether the claims can cover the complex metadata signals used in modern advertising and content platforms.
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: The specification describes tags broadly, including objective descriptors ("Chinese," "rice"), subjective or semi-subjective terms ("tasty," "healthy"), and even "tangential tags" that are only relevant when correlated with other information ("hipster," "yolo") ('705 Patent, col. 9:4-10, col. 9:53-66). This could support a construction that covers a wide variety of metadata, keywords, or other signals.
      • Evidence for a Narrower Interpretation: The primary embodiments and figures consistently depict tags as explicit, human-readable text labels associated with a static image of a physical object (e.g., '705 Patent, Figs. 4D, 4E). A defendant could argue this context limits "tags" to explicit, user-facing labels rather than internal, algorithmic signals or vectors.

  • The Term: "a next image...representing a different physical object" ('786 Patent, Abstract; '705 Patent, col. 2:18-20)

    • Context and Importance: This term is crucial for establishing the iterative nature of the claimed invention. The infringement analysis must show that the accused systems present a sequence of distinct items.
    • Intrinsic Evidence for Interpretation:
      • Evidence for a Broader Interpretation: The specification's general description of the problem and solution focuses on guiding a user through a series of choices. This could support reading the term on any sequence of distinct content, such as a series of different recommended videos or different advertisements.
      • Evidence for a Narrower Interpretation: The patent repeatedly uses the example of "food dishes" and other tangible "physical objects" ('705 Patent, col. 1:62, col. 9:1-4). This may support an argument that the claims are limited to tangible goods and do not extend to digital content like videos or services offered in advertisements, which may not be considered "physical objects" in the same sense.

VI. Other Allegations

  • Indirect Infringement: The complaint does not contain allegations of indirect infringement (induced or contributory). Both counts allege direct infringement under 35 U.S.C. § 271(a) (Compl. ¶¶18, 25).
  • Willful Infringement: The complaint does not use the word "willful," but it lays a foundation for such a claim. It alleges that Defendant "made no attempt to design around the claims" and "did not have a reasonable basis for believing that the claims of the ['786 and '705 Patents] were invalid" (Compl. ¶¶19-20, 26-27). Furthermore, the prayer for relief requests a judgment that this is an "exceptional case" under 35 U.S.C. § 285, which is the statutory basis for awarding enhanced damages and attorney's fees, often in cases of willful infringement (Compl. Prayer for Relief ¶C).

VII. Analyst’s Conclusion: Key Questions for the Case

This case appears to present two primary challenges for the court and the parties:

  • A core issue will be one of definitional scope and claim construction: Can terms like "tags" and "session," rooted in the patent's specific examples of a "Tinder-for-food" application, be construed broadly enough to encompass the fundamentally different and more complex technical environments of the YouTube content recommendation engine and the Google Ads auction and delivery system?
  • A key challenge will be one of evidentiary proof: The complaint makes conclusory allegations of infringement while referring to non-public exhibits. The case will likely depend on whether Plaintiff can, through discovery into Google's highly proprietary systems, unearth specific evidence that these complex machine-learning-based platforms in fact perform the discrete, rule-based steps recited in the patent claims, such as explicitly "removing tags" or "adjusting weights" in the manner claimed.