DCT
1:24-cv-00552
Sterling Computers Corp v. X Corp
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Sterling Computers Corporation (California)
- Defendant: X CORP. (Nevada)
- Plaintiff’s Counsel: Padmanabhan & Dawson, P.L.L.C.
- Case Identification: Sterling Computers Corporation v. X CORP., 1:24-cv-00552, W.D. Tex., 05/22/2024
- Venue Allegations: Plaintiff alleges venue is proper because Defendant has a regular and established place of business in the district, specifically an office in Austin, Texas.
- Core Dispute: Plaintiff alleges that Defendant’s social media platform, X (formerly Twitter), infringes a patent related to methods for determining the relevance of electronic content based on user behavior.
- Technical Context: The technology addresses the challenge of personalizing and ranking content in high-volume electronic messaging environments, a fundamental component of modern social media feed algorithms.
- Key Procedural History: The patent-in-suit claims priority to a provisional application filed in 2006. The complaint does not mention any other prior litigation, licensing history, or post-grant proceedings involving the patent.
Case Timeline
| Date | Event |
|---|---|
| 2006-01-13 | '217 Patent Priority Date (Provisional Application) |
| 2010-05-11 | U.S. Patent No. 7,716,217 Issues |
| 2024-05-22 | Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 7,716,217 - "Determining Relevance of Electronic Content"
- Patent Identification: U.S. Patent No. 7,716,217, "Determining Relevance of Electronic Content", issued May 11, 2010.
The Invention Explained
- Problem Addressed: The patent addresses the problem of users being inundated with "unwieldy volumes of e-mail messages," making it difficult to determine which content is most important ('217 Patent, col. 1:23-34; Compl. ¶9). Prior art solutions like spam filters were deemed insufficient because they only removed unsolicited messages, rather than helping to prioritize relevant ones among legitimate communications (Compl. ¶10).
- The Patented Solution: The invention is a system that learns and scores the relevance of electronic content for a specific user by monitoring that user's actions (e.g., reading, replying) and the actions of other, similar ("cognate") users ('217 Patent, col. 2:47-66). A relevance analysis engine combines these behavioral inputs to calculate a relevance score, moving beyond simple keyword filtering to a personalized, dynamic ranking system (Compl. ¶¶11-13; ’217 Patent, col. 6:1-12).
- Technical Importance: This approach represented a shift from generalized content filtering to user-specific, behavior-based relevance scoring, a foundational concept for personalizing information feeds in high-volume digital environments ('217 Patent, col. 1:23-44).
Key Claims at a Glance
- The complaint asserts independent claim 1 ('217 Patent, col. 47:25-48:2).
- The essential elements of independent claim 1 include:
- A computer-implemented system for determining a relevance score for a piece of electronic content.
- The system includes a computer processor and computer-readable storage medium with program modules.
- A "monitoring module" to track actions by a plurality of users.
- A "first relevance measurement module" to determine a first measure of relevance based on the actions of the user.
- A "second relevance measurement module" configured to identify a "cognate" user and determine a second measure of relevance based on factors including the sender's local and global importance, content relevance, and the actions of the cognate user.
- A "relevance analysis module" to determine the final relevance score based on the first and second measures of relevance.
III. The Accused Instrumentality
Product Identification
- The social media platform known as "Twitter," or "X" (Compl. ¶15).
Functionality and Market Context
- The complaint alleges that the accused instrumentality is Twitter's "recommendation algorithm," which is used to select and rank content for a user's "For You" timeline (Compl. ¶31, p. 5).
- This system is described as processing approximately 500 million daily Tweets and using a machine learning model to "predict the relevance of each candidate Tweet" (Compl. ¶31, p. 6).
- The system's functionality relies on "a set of core models and features that extract latent information from Tweet, user, and engagement data" to personalize the user's feed (Compl. ¶31, p. 7).
IV. Analysis of Infringement Allegations
A flowchart from a blog post illustrates the accused system's major components, including modules for CANDIDATE SOURCES, a HOME MIXER, and a HEAVY RANKER used to construct a user's timeline (Compl. ¶31, p. 7).
’217 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| A computer-implemented system for determining a relevance score of a piece of electronic content sent from a sender to a user... | Twitter/X is a computer-implemented system that determines a relevance score for Tweets, described as using a "recommendation algorithm" to "give each Tweet a score" for ranking in the "For You timeline." | ¶31 | col. 2:47-56 |
| a monitoring module configured to track actions by a plurality of users associated with an application for managing electronic content; | The Twitter system's foundation is a set of models that "extract latent information from Tweet, user, and engagement data," which allegedly constitutes tracking user actions. | ¶31 | col. 6:40-44 |
| a first relevance measurement module... configured to determine a first measure of relevance... based at least in part on one or more actions of the user on the piece of electronic content; | The Twitter system applies "heuristics and filters, such as filtering out Tweets from users you’ve blocked... and Tweets you’ve already seen," which are alleged to be user actions that inform relevance. | ¶31 | col. 8:35-39 |
| a second relevance measurement module... configured to... identify at least one other... user that is... cognate to the user, and determine a second measure of relevance... | The system analyzes a "Social Graph" to find what is relevant based on "the engagements of people you follow or those with similar interests." The second measure of relevance is allegedly derived from factors like "In-Network Source" (local importance), "SimClusters" (global importance), and "content similarity" (content relevance). | ¶31 | col. 14:60-67 |
| a relevance analysis module configured to determine the relevance score... based at least in part on the first determined measure of relevance and on the second determined measure of relevance. | The Twitter system's "Ranking" stage allegedly combines various features and data inputs to "predict the relevance of each candidate Tweet," which is the primary signal for ranking. This is alleged to be the function of the relevance analysis module. | ¶31 | col. 6:44-54 |
Identified Points of Contention
- Scope Questions: The patent specification is heavily focused on the context of corporate "e-mail" systems ('217 Patent, col. 1:12). A central dispute may be whether the term "electronic content" should be interpreted as limited to that e-mail context or if it is broad enough to encompass short-form social media posts ("Tweets") on a public platform.
- Technical Questions: The complaint maps features of Twitter's algorithm, such as "Social Graph" and "SimClusters," to the claimed "modules." A technical question is whether Twitter’s system, described as a complex machine learning pipeline, actually contains distinct software structures corresponding to the claimed "first" and "second" relevance measurement modules, or if these functions are inseparably integrated in a way that does not meet the claim's architectural requirements.
V. Key Claim Terms for Construction
- The Term: "cognate to the user"
- Context and Importance: This term is critical to the scope of the "second relevance measurement module." Infringement of this element hinges on whether the users that Twitter's algorithm analyzes (e.g., users one follows, users with similar interests) fall within the patent's definition of "cognate." Practitioners may focus on this term because its construction will determine whether loose social-graph connections qualify as the more structured relationships implied by the patent's examples.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification defines a cognate group based on "similar past user actions and assessments" ('217 Patent, col. 14:65-67), which could support a broad reading that includes users with algorithmically determined similar interests.
- Evidence for a Narrower Interpretation: The specification’s examples of cognate entities often refer to formal organizational structures, such as a "human resources employee" or a "supervisor" ('217 Patent, col. 14:50-54). This language could support an argument that "cognate" requires a more defined relationship than simply following another user or sharing interests on a social media platform.
VI. Other Allegations
The complaint alleges only direct infringement and does not contain allegations of indirect or willful infringement.
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of contextual scope: can the claims, which are rooted in the patent’s disclosure of a system for prioritizing corporate "e-mail", be construed broadly enough to cover the ranking of "Tweets" in a public social media feed?
- A key evidentiary question will be one of structural correspondence: does the accused Twitter algorithm, a complex machine learning system, possess the distinct "first" and "second relevance measurement modules" recited in Claim 1, or is there a fundamental architectural difference between the claimed invention and the accused system?
- The infringement analysis will likely turn on a definitional question: can the term "cognate to the user", which the patent illustrates with examples of organizational relationships, be interpreted to read on the algorithmically-defined and user-selected affiliations (e.g., "people you follow," "similar interests") within the Twitter social graph?