PTAB

IPR2021-00797

accessiBe Ltd v. AudioEye Inc

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Modular Systems and Methods for Selectively Enabling Cloud-Based Assistive Technologies
  • Brief Description: The ’934 patent relates to systems for automatically remediating websites to comply with accessibility guidelines. The disclosed methods involve accessing a website's code via its Document Object Model (DOM), identifying elements that lack meaningful textual descriptions, and programmatically inserting appropriate text for use by assistive technologies like screen readers.

3. Grounds for Unpatentability

Ground 1: Claims 1-10 and 15-16 are obvious over Springer, Bigham, and W3C DOM

  • Prior Art Relied Upon: Springer (Application # 2004/0148568), Bigham (a 2007 conference paper), and W3C DOM (a 1998 W3C Recommendation).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Springer taught a system of "checker" and "fixer" modules to automatically find and remediate web accessibility issues. Specifically, its "ImageChecker" and "ImageMapChecker" tools crawl a website's "document tree" (a DOM) to detect images and image links that lack valid "alt" attributes. Once an issue is detected, a corresponding "fixer" module remediates it. Petitioner asserted that while Springer identified the problem, Bigham disclosed specific, automated solutions for generating the missing alt text for such hyperlinked elements. Bigham taught a "link-based labeling method" that follows a hyperlink to analyze the context of the destination page and also taught using heuristic analysis on an image's filename to produce meaningful alt text. The combination of Springer's automated remediation framework with Bigham's specific alt-text generation techniques allegedly taught the core limitations of independent claim 1.
    • Motivation to Combine: A POSITA would combine Bigham's specific filename and link-based analysis methods into Springer's general framework to create a more robust, fully automated accessibility solution. Bigham provided a simple and tested method for determining appropriate alt text, which would improve Springer’s system by removing the need for potential operator/user input. Petitioner further contended it would be obvious to use the well-known W3C DOM standard as the "document object model" for the combined system, as it was the dominant format for programmatically modifying website code in the early 2000s and is functionally what Springer’s "document tree" represents.
    • Expectation of Success: Because both Springer and Bigham operate in the web accessibility field and address the same problem of remediating untagged hyperlinked elements, a POSITA would have reasonably expected the combination to work as intended.

Ground 2: Claims 11-14 and 17-21 are obvious over Springer, W3C DOM, Nandakumar, and Dove

  • Prior Art Relied Upon: Springer (Application # 2004/0148568), W3C DOM (a 1998 W3C Recommendation), Nandakumar (Patent 10,394,579), and Dove (a 2015 online article).

  • Core Argument for this Ground:

    • Prior Art Mapping: This ground addresses claim 11, which recites using AI image recognition to generate alt text. Petitioner argued that Springer’s "ImageFixer" tool provided a framework for adding alt text to images but did not limit how that text is determined, creating a clear opening for a more advanced method. Nandakumar and Dove supplied this missing element. Nandakumar taught a system that, upon identifying an inaccessible image, forwards the image to an "image-based content retrieval (IBCR) application programming interface (API)" which returns a "text content description relevant to that particular image." Dove described a real-world, publicly available API (CloudSight) that performed this exact function using deep learning to analyze images and return textual descriptions.
    • Motivation to Combine: A POSITA would have been motivated to incorporate the superior AI-based image recognition taught by Nandakumar and Dove into Springer's remediation system. This would replace less reliable methods or manual input with a more powerful, automated, and readily available technology for generating accurate alt text for images. The AI services described in the references were expressly designed as APIs for easy integration into a variety of applications, making them an obvious and desirable enhancement for the Springer system.
    • Expectation of Success: Given that the AI services in Nandakumar and Dove were designed as public APIs for easy integration, a POSITA would have had a high expectation of success in modifying Springer's ImageFixer tool to call such an API to determine alt text for untagged images with minimal modification.

  • Additional Grounds: Petitioner asserted additional obviousness challenges substituting Lehota (Application # 2010/0205523) for Springer. These grounds argued Lehota's system, which uses JavaScript widgets to modify a website's DOM to fix accessibility issues, would have been obviously combined with Bigham for hyperlink analysis (Ground 3) and with Nandakumar and Dove for AI image recognition (Ground 4) based on similar design modification theories.

4. Arguments Regarding Discretionary Denial

  • Petitioner argued that discretionary denial under 35 U.S.C. §314(a) based on Fintiv factors was unwarranted. It was contended that the parallel district court litigation was in a very early stage, with no significant investment made by the court and no substantive proceedings (like claim construction) having occurred. The petitioner noted the tentatively scheduled trial date was approximately 17 months away, providing ample time for a Final Written Decision before trial. Furthermore, the petition challenged claims (16 and 19) not asserted in the litigation and presented different prior art and arguments, reducing the overlap between the proceedings. Petitioner asserted these factors, combined with the strong merits of the unpatentability arguments, weighed heavily against discretionary denial.

5. Relief Requested

  • Petitioner requested institution of an inter partes review and cancellation of claims 1-21 of Patent 10,444,934 as unpatentable under 35 U.S.C. §103.