Obviousness

Obviousness Analysis under 35 U.S.C. § 103 for US Patent 11875275

This analysis identifies combinations of prior art references and common general knowledge that would render the independent claims of US Patent 11875275 obvious to a person having ordinary skill in the art (PHOSITA) before the patent's priority date of October 17, 2019. The core invention of US11875275 is facilitating the creation and execution of custom time series models in computer analytics systems using a "model creation wizard interface" to receive user-defined parameters.

Prior Art References Considered

1. US 9,130,971 B2 (Site-Based Search Affinity), issued September 8, 2015: This patent, frequently referenced within the text of US11875275, describes a robust computer analytics system (e.g., the SPLUNK® ENTERPRISE system) designed for collecting, indexing, and searching massive volumes of machine data. It handles events associated with specific points in time ("time series data"), uses flexible schemas for on-the-fly information extraction, and supports functionalities like anomaly detection and prediction of future outcomes based on data analysis. The patent explicitly notes that "models are executed by computers to analyze the data and detect information about the users and hardware that directly or indirectly cause the data," and that "anomaly detection is used for systems to detect fraud, security threats, and the existence of faulty data." While it broadly refers to "models" for these purposes, it also states that "creating a model is generally performed by enterprise software providers. When the enterprise software is provided to the customers or used by customers, the customers are often only able to use a fixed set of models created by the enterprise software provider."

2. Common General Knowledge (CGK) of User Interface Design for Complex Software (pre-Oct 2019): Prior to October 2019, "wizard interfaces" were a well-established and ubiquitous design pattern in software development for guiding users through multi-step processes, configurations, or creations that required inputting various parameters. This was a common approach to simplify complex tasks for users who might not have expert-level technical knowledge.

3. Common General Knowledge (CGK) of Analytical Software and Custom Model Creation (pre-Oct 2019): Before October 2019, various analytical software tools (e.g., Microsoft Excel with its Data Analysis ToolPak, statistical packages like SPSS or SAS, or open-source environments like R and Python with libraries such as scikit-learn or statsmodels) routinely provided users with interfaces to define parameters for diverse statistical, machine learning, and time series models. These tools allowed users to specify inputs, select model types, and define specific parameters (e.g., window sizes, thresholds, regression coefficients) to generate results that identified properties not explicitly present in the raw input data (e.g., trends, correlations, anomalies, predictions).

Obviousness Argument

The independent claims of US Patent 11875275 (Claims 1, 9, and 17) cover a method, system, and non-transitory computer-readable medium, respectively, for:

1. Receiving user-defined parameters for a custom model via a model creation wizard interface.
2. Transforming the custom model definition into an executable custom model using these parameters.
3. Executing the custom model to identify properties in data not explicitly present.

Combination: US 9,130,971 B2 in view of Common General Knowledge of UI Wizards and Analytical Software

A PHOSITA, at the time of the invention, would have found the claimed invention obvious by combining the teachings of US 9,130,971 B2 with common general knowledge in software engineering and data analytics.

Motivation to Combine:
The US 9,130,971 B2 patent clearly describes sophisticated computer analytics systems that ingest and process massive amounts of machine data, enabling searches and supporting high-level analytical tasks like anomaly detection and predicting future outcomes. However, it identifies a limitation where "creating a model is generally performed by enterprise software providers" and customers "are often only able to use a fixed set of models." A PHOSITA would recognize this as a significant impediment to users needing specialized or custom analyses for their unique operational contexts. There would be a strong motivation to enhance such a powerful data analytics platform by enabling end-users to create their own custom analytical models. This would empower a broader user base to extract more specific and relevant insights from their machine data without requiring extensive programming or data science expertise for every custom scenario.

Application to Claim Elements:

• "receiving, by a computer system, one or more user-defined parameters of a custom model definition via a model creation wizard interface;"

  • US 9,130,971 B2 discloses a "computer system" for data intake and query, processing machine data and events, and capable of applying "models" for analysis like anomaly detection.
  • CGK of UI Wizards: The idea of using a "wizard interface" to gather "user-defined parameters" for a software function was well-known and routinely applied to simplify complex configuration or creation processes in enterprise software, including those related to data analysis, reporting, and system setup. Extending this established UI paradigm to define parameters for custom analytical models would be a straightforward application of known techniques to address the identified problem of limited user customization.
  • CGK of Analytical Software: Existing analytical software commonly provided interfaces (some wizard-like) for users to input parameters for various models. Therefore, incorporating such a user-friendly parameter input mechanism into an enterprise analytics system would be an obvious design choice for improving usability and customizability.

• "transforming, by the computer system, the custom model definition into a custom model using the one or more user-defined parameters;"

  • US 9,130,971 B2 describes models used in the system, implying the system's ability to execute them. The patent further explains that "Programming is used to transform the structure into a workable model that a computer can interpret."
  • CGK of Analytical Software and Software Engineering: The step of taking user-defined parameters (e.g., statistical coefficients, machine learning algorithm choices, time-window settings) and configuring or generating executable code or logic to create a functional "custom model" is a fundamental and inherent aspect of any configurable analytical software. A PHOSITA would understand that once parameters are received, the system must process them to create an operable model instance. Modern analytics platforms frequently employed techniques like dynamic code generation, script execution, or API calls to underlying analytical engines based on user inputs.

• "executing, by the computer system, the custom model to identify one or more properties in data that are not explicitly present in the data."

  • US 9,130,971 B2 explicitly states that its models "detect information about the users and hardware that directly or indirectly cause the data," and that "the properties detected are not explicitly in the input data." It specifically mentions "anomaly detection" and "predict future outcomes" as examples of what models do.
  • CGK of Analytical Software: The primary purpose of any analytical model (statistical, machine learning, time series) is to derive insights, predictions, or classifications (i.e., "properties") that are not immediately obvious or "explicitly present" in the raw input data. Therefore, executing a custom model to achieve this goal is the expected and inherent function of such a model within any analytics system.

Conclusion

Based on the foregoing, a PHOSITA would have been motivated to combine the established data analytics capabilities of a system like that described in US 9,130,971 B2 with the common general knowledge of using wizard interfaces for simplified parameter input and the known functionalities of analytical software to define and execute custom models. This combination would lead to the obvious development of a system and method that allows users to define custom time series models through a wizard, transform these definitions into executable models, and then run them to identify implicit properties in data, thereby overcoming the acknowledged limitation of fixed enterprise-defined models. The claimed invention, therefore, would have been obvious under 35 U.S.C. § 103.