DCT
1:24-cv-00123
Nielsen Co US LLC v. VideoAmp Inc
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: The Nielsen Company (US), LLC (Delaware)
- Defendant: VideoAmp, Inc. (Delaware)
- Plaintiff’s Counsel: Potter Anderson & Corroon LLP
- Case Identification: The Nielsen Company (US), LLC v. VideoAmp, Inc., 1:24-cv-00123, D. Del., 01/31/2024
- Venue Allegations: Venue is alleged to be proper in the District of Delaware because the Defendant is a Delaware corporation and therefore resides in the district.
- Core Dispute: Plaintiff alleges that Defendant’s audience measurement platform infringes two patents related to methods for correcting inaccurate television viewing data by modeling high-quality panel data to refine large-scale set-top box data.
- Technical Context: The technology addresses the challenge of accurately measuring media consumption, a critical function for the multi-billion dollar television advertising market, by correcting for "phantom viewing" where a device is on but not being watched.
- Key Procedural History: The complaint notes that the two patents-in-suit descend from the same parent application and share a "Common Specification." This shared disclosure will be a central piece of intrinsic evidence for interpreting claim terms across both patents.
A. Case Timeline
| Date | Event |
|---|---|
| 2015-10-08 | Priority Date for ’058 and ’250 Patents |
| 2018-07-18 | VideoAmp acquires IronGrid to process set-top box data |
| 2023-12-26 | U.S. Patent No. 11,856,250 Issues |
| 2024-01-09 | U.S. Patent No. 11,871,058 Issues |
| 2024-01-31 | Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
A. U.S. Patent No. 11,871,058 - "Methods and Apparatus to Determine a Duration of Media Presentation Based on Tuning Session Duration"
1. The Invention Explained
- Problem Addressed: The patent addresses the inaccuracy of "tuning data" collected from devices like set-top boxes (STBs) (Compl. ¶25). This data can indicate that a device is tuned to a channel, but it cannot determine if the associated television is actually on, or if it is displaying content from a different source (e.g., a Blu-Ray player), a phenomenon referred to as "phantom viewing" (’058 Patent, col. 2:58-3:14; Compl. ¶25).
- The Patented Solution: The invention solves this problem by using two different sets of households. For a "first subset" of households where high-quality "presentation session data" (e.g., from a metered panelist) is available, the system builds a statistical model that correlates tuning session durations with actual viewing durations (Compl. ¶¶26, 30). This model is then applied to a "second subset" of households, for which only the less-reliable tuning data is available, to estimate the actual, or "expected," duration of media presentation (’058 Patent, col. 4:4-10; Abstract).
- Technical Importance: This approach allows for the correction and enhancement of massive but unreliable "big data" sets (from STBs) by leveraging smaller, but more accurate, "truth sets" (from metered panels), thereby improving the overall accuracy and value of audience measurement analytics (Compl. ¶¶15, 40).
2. Key Claims at a Glance
- The complaint asserts independent claim 1 and dependent claims 2-5, 9, 12-19, and 22 (Compl. ¶64).
- The essential elements of independent claim 1 (a method claim) include:
- Obtaining tuning session data from a plurality of media presentation devices, where presentation data is available for a first subset of environments but not for a second subset.
- For the first subset, obtaining presentation session data indicating the actual duration media was presented.
- Generating a model that relates the tuning session durations to the corresponding presentation session durations from the first subset.
- Using the model to determine an "expected presentation session duration" for a given tuning session in the second subset.
- Causing the presentation of the determined expected duration.
- The complaint reserves the right to assert other claims (Compl. ¶64).
B. U.S. Patent No. 11,856,250 - "Methods and Apparatus to Determine a Duration of Media Presentation Based on Tuning Session Duration"
1. The Invention Explained
- Problem Addressed: As a continuation of the same patent family with a "Common Specification," the ’250 Patent addresses the identical problem of "phantom viewing," where raw STB tuning data does not accurately reflect actual media consumption by a viewer (Compl. ¶¶23, 25).
- The Patented Solution: The patented solution is a system that implements the same core process as the ’058 Patent. The system accesses both tuning data and more accurate presentation data from a "first subset" of media environments to generate a model, which is then used to estimate presentation sessions for a "second subset" of environments that only provide tuning data (’250 Patent, Abstract; Compl. ¶26).
- Technical Importance: The technical importance is identical to that of the ’058 Patent, enabling the use of large-scale STB data by correcting its inherent inaccuracies with data from more reliable sources (Compl. ¶40).
2. Key Claims at a Glance
- The complaint asserts independent claim 1 and dependent claims 2-5, 9, 11, 14-19, and 22 (Compl. ¶71).
- The essential elements of independent claim 1 (a system claim) include:
- A computing system with processor circuitry and instructions to perform steps.
- Accessing tuning session data from a plurality of devices, separating environments into a first subset (with presentation data) and a second subset (without).
- Accessing the presentation session data for the first subset.
- Generating a model relating tuning session durations to presentation session durations from the first subset.
- Using the model to determine an expected presentation session duration for a given tuning session in the second subset.
- Outputting the determined expected duration.
- The complaint reserves the right to assert other claims (Compl. ¶71).
III. The Accused Instrumentality
- Product Identification: The accused instrumentality is the "VideoAmp Platform," which includes the "Infringing Apparatus" and "Infringing Method" used to provide audience measurement products and services (Compl. ¶59).
- Functionality and Market Context:
- The VideoAmp Platform is alleged to create a "unified TV data set" by "commingling" large-scale Set-Top Box (STB) data with Automatic Content Recognition (ACR) data from Smart TVs (Compl. ¶¶49, 51). The complaint alleges that this process is used specifically to correct for inaccuracies in STB data, such as "phantom viewing," where a set-top box is on but the television is off (Compl. ¶¶53, 56).
- The complaint alleges, based on a CIMM Whitepaper, that VideoAmp’s process involves using "tuning data from matched homes with both Set Top Box and Smart TV ACR device matches to inform calibration of combined data set, including un-matched homes" (Compl. ¶56). A diagram in the complaint shows a five-step process that includes using ACR data to "inform algorithms that classify zombie viewing in the STB data feed" (Compl. p. 22). This diagram illustrates a five-step process for commingling STB and ACR data, including a step to use ACR data to inform algorithms for classifying "zombie viewing" (Compl. p. 22).
- VideoAmp is positioned as a direct competitor and "primary rival to Nielsen," offering its platform as an "alternative" currency for buying and selling television advertising (Compl. ¶¶16, 60).
IV. Analysis of Infringement Allegations
A. '058 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| obtaining respective tuning session data from a plurality of media presentation devices... wherein presentation session data is available for a first subset... and... not available for a second subset | The VideoAmp Platform allegedly ingests STB data and uses "matched homes" (with STB and ACR data) to calibrate data for "un-matched homes" (with only STB data). | ¶¶50, 56 | col. 25:56-26:8 |
| obtaining, for ones of the first subset... respective presentation session data indicative of a presentation session duration | The platform allegedly ingests Smart TV ACR data, which provides more accurate viewership information corresponding to the "presentation session data" from the "first subset." | ¶¶50, 51 | col. 26:9-14 |
| generating... a model relating the respective tuning session durations to corresponding presentation session durations from ones of the first subset | VideoAmp allegedly uses data from "matched homes" to "inform calibration" and applies algorithms to create a "Set On/Set Off model" to correct for phantom viewing. | ¶¶56, p. 22 | col. 26:15-18 |
| based on the model, determine... an expected presentation session duration for a given tuning session | The platform allegedly applies the generated model/calibration to the STB data from "un-matched homes" to correct for inaccuracies like a TV being off. | ¶56 | col. 26:18-22 |
| causing... presentation of the expected presentation session duration | VideoAmp allegedly reports the results of its commingling process, which are used by clients as an advertising "currency." | ¶59 | col. 26:22-25 |
- Identified Points of Contention:
- Scope Questions: A central question will be whether VideoAmp's "ACR data" from smart TVs falls within the scope of the term "presentation session data" as used in the patent, which provides Local People Meters (LPMs) as a primary example. Likewise, does VideoAmp's use of "matched homes" and "un-matched homes" map directly onto the patent's "first subset" and "second subset" of media presentation environments?
- Technical Questions: What evidence demonstrates that VideoAmp’s alleged "calibration" process and algorithms for classifying "zombie viewing" perform the specific claimed step of "generating a model relating... tuning session durations to corresponding presentation session durations"? The analysis will likely focus on whether the accused functionality is structurally and functionally equivalent to this claimed modeling step.
B. '250 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| access respective tuning session data... wherein presentation session data is available for a first subset... and... not available for a second subset | The VideoAmp Platform allegedly accesses STB data and differentiates between "matched" data sets (first subset) and "un-matched" homes (second subset). | ¶¶50, 56 | col. 25:51-26:7 |
| access, for ones of the first subset... respective presentation session data | The platform allegedly accesses Smart TV ACR data, which corresponds to the patent's "presentation session data." | ¶¶50, 51 | col. 26:8-12 |
| generate a model relating the respective tuning session durations to corresponding presentation session durations | VideoAmp allegedly generates a "Set On/Set Off model" based on the "matched" STB and ACR data to "inform calibration" for other datasets. | ¶¶56, p. 22 | col. 26:13-16 |
| based on the model, determine... an expected presentation session duration | The platform allegedly applies this model to STB data from "un-matched homes" to estimate actual viewing and correct inaccuracies. | ¶56 | col. 26:16-20 |
| output the expected presentation session duration | VideoAmp allegedly provides the resulting corrected data to its clients for use in advertising transactions. | ¶59 | col. 26:20-22 |
- Identified Points of Contention: The points of contention for the ’250 Patent are identical to those for the ’058 Patent, as the core technology and infringement theories are the same. The dispute will center on the definitional scope of key terms like "presentation session data" and the functional equivalence between VideoAmp’s "commingling" process and the patent's specific claimed modeling method.
V. Key Claim Terms for Construction
- The Term: "presentation session data"
- Context and Importance: This term is the lynchpin of the infringement allegation. The plaintiff's case appears to rely on construing this term to read on the "ACR data" that VideoAmp allegedly uses. The defendant may argue for a narrower construction limited to the patent's specific examples.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification defines the data functionally as capturing "when the television is on and displaying media from the reported content source" and being more accurate than tuning data (Compl. ¶26). The patent also describes LPMs as an "example" source, which may suggest the term is not limited to that specific embodiment (’058 Patent, col. 4:41-50).
- Evidence for a Narrower Interpretation: The specification repeatedly and consistently uses data from "LPMs" (Local People Meters) as the source of the "metering data" used to build the claimed models (’250 Patent, col. 4:4-10). A defendant could argue this consistent focus on a specific type of panel-based data limits the scope of the claims to models built from that data type, not from large-scale ACR data.
- The Term: "generate a model relating the respective tuning session durations to corresponding presentation session durations"
- Context and Importance: This claim element describes the specific inventive step. Practitioners may focus on this term because the infringement case depends on whether VideoAmp's alleged "calibration," "adjustments," and "algorithms" (Compl. ¶56) are the same as or equivalent to generating this specific type of model.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The claim language itself is functional, requiring a "model relating" the two types of durations. The abstract also broadly describes the invention as estimating presentation duration "based on the model" (’250 Patent, Abstract). This may support a construction that covers any process that uses one duration type to predict the other.
- Evidence for a Narrower Interpretation: The specification provides detailed examples of the models, including frequency distributions, conditional distributions, and cumulative distributions based on time durations (e.g., ’250 Patent, FIGS. 8A-8E). A defendant may argue that these specific embodiments define the metes and bounds of the claimed "model," and that a different type of algorithm (e.g., a binary classifier for "zombie viewing") does not meet this limitation.
VI. Other Allegations
- Willful Infringement: The complaint alleges that VideoAmp has knowledge of the patents-in-suit "as of the service date of the Complaint" and that its infringement is willful and deliberate from that date forward (Compl. ¶¶67, 74). The complaint does not allege any pre-suit knowledge of the patents.
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of definitional scope: can the term "presentation session data", which the patents primarily illustrate using data from panelist-based Local People Meters (LPMs), be construed to cover the large-scale Automatic Content Recognition (ACR) data that the complaint alleges VideoAmp uses to correct its viewership metrics?
- A key evidentiary question will be one of functional equivalence: does VideoAmp’s alleged process of using a "matched data set" of STB and ACR data to "inform calibration" and "classify zombie viewing" for other data sets perform the specific, claimed step of "generating a model relating... tuning session durations to corresponding presentation session durations," or is there a fundamental mismatch in the technical operation of the two processes?