Holding: A patent that claims the application of machine learning to a new data environment without disclosing improvements to the applied machine learning model is directed to an abstract idea and is patent ineligible under Section 101.
Recentive sued Fox for the infringement of US Patents No. 10,911,811; 10,958,957; 11,386,367; and 11,537,960. The ‘367 Patent and the ‘960 Patent are directed to the scheduling of live events and teach the use of a machine learning model that is trained using a set of training data to create an optimized television schedule. The ’811 Patent and the ‘957 Patent are directed to network maps that determine the programs that are displayed on a broadcaster’s channels in certain geographic markets at particular times. The network maps are created using training data in conjunction with a machine learning model. A user may input target features, such as ratings for a particular market or particular market, to achieve a desired result.
Fox alleged that the Recentive patents were ineligible under Section 101. In response, Recentive acknowledged that the idea of preparing network maps already existed and were typically prepared by human beings. The claims of the Recentive patents did not claim the machine learning technique itself, but instead claimed the application of machine learning techniques to the specific context of television event scheduling and network map creation. Recentive argued that their patents claimed a unique application of machine learning to create customized algorithms to automatically create schedules and maps using iterative training for different event parameters and target features.
The district court found the Recentive patents to be directed to ineligible subject matter using the two-step inquiry of Alice. The district court found the claims to be directed to the abstract ideas of producing network maps and event schedules using known, generic mathematical techniques. The district court also found that the claims were not directed to an inventive concept that amounts to significantly more than the ineligible concept, since the claims described well-known machine learning techniques.
The Federal Circuit reviewed the district court’s decision in view of the Alice inquiry, and upheld the district court’s finding that the asserted claims of the Recentive patents are patent ineligible. In the Step One of the Alice analysis, for software patents, the Federal Circuit asked whether the claims focus on an asserted improvement in computer capabilities or if the claims focus on a process that qualifies as an abstract idea for which computers a merely a tool. Since Recentive conceded that they were not claiming the machine learning technique itself, the claims of the Recentive patents were directed to ineligible, abstract subject matter. The claims of the Recentive patents used generic, conventional machine learning technology to carry out methods for generating event schedules and network maps. The fact that the claims suggest iterative training or dynamic adjustment to the machine learning techniques do not represent a technological improvement. These are basic features of any machine learning process. Furthermore, the claims failed to describe the steps through which the machine learning technique enhanced network mapping or event scheduling.
The Recentive patents simply describe the use of machine learning in the new environment of network mapping or event scheduling. Before machine learning, both of these activities were performed by humans using similar parameters. The fact that machine learning is being used in a new field does not make these patents patent eligible, nor does the application of existing machine learning technology to a new database provide eligibility. Patents may be directed to an abstract idea when they disclose the use of an already existing technology and its basic functions as a tool to execute a claimed process.
Additionally, the Federal Circuit found that claimed methods do not become patent eligible merely because they improve speed and efficiency of a task that was formerly performed by humans. Increased speed and efficiency due to the use of computers, without improved function of the computer technique, does not create patent eligibility.
In Step Two of the Alice analysis, the Federal Circuit considered whether the elements of the claims transform the nature of the claim into a patent eligible application. This requires an inventive concept that is more than simply claiming the abstract idea and stating to “apply” this idea. Recentive argued that the inventive concept in its claims was using machine learning to dynamically generate optimized network maps and schedules based on real time data and updating those maps and schedules based on changing conditions. However, the Federal Circuit found that this concept is little more than the abstract idea itself and fails to identify anything that would transform the abstract idea into a patent eligible application.
Federal Circuit Precedential 
Leave a comment