If you draft patent applications directed to algorithms, chances are you will eventually deal with rejections under Section 112 where the examiner alleges that the specification does not have sufficient disclosure of a detailed algorithm. This requirement is summarized by the USPTO’s guidance on written description (functional claiming) from 2019 (here).
While there is much to discuss regarding these guidelines, today we focus on the specific issue where the patent application is drafted to exactly support/match the claimed elements. This is a common approach for some applicants where the claimed actions/elements of the inventive algorithm are then used as the skeleton of the patent application, without alternative descriptive language. Some applicant consider this the pinnacle of patent drafting due the tight correlation between specification and claims. However, it can lead to situations that can be difficult to overcome in prosecution. One such example is illustrated by an Adobe application (14/923,153).
The invention here relates planning and managing advertising campaigns that include both digital media advertising and TV advertising, such as a cross format advertising campaign which targets advertisements for both television and online viewing. The invention provides a planning tool for an advertising campaign planner to determine the optimal advertising budget allocation between online and television advertisements for a particular advertising campaign strategy.
Claim 1:
1. In a digital medium environment for real-time bidding and selection of advertising opportunities corresponding to viewers simultaneously accessing websites via computing devices, a computerized method for generating and executing advertising campaigns of de-duplicated reach based on television and online viewership data, the method comprising:
receiving, via one or more processors, a fused data set of viewership data comprising data related to television viewing and digital media viewing of advertisements, wherein, for a plurality of viewers categorized as watching advertisements, the fused data set comprises digital media viewership data for the plurality of viewers;
receiving a list of a plurality of television advertisements related to a client;
identifying one or more viewers of the plurality of viewers of the fused data set categorized as not watching a respective television advertisement of the plurality of television advertisements;
identifying media properties comprising URLs visited by the identified one or more viewers of the plurality of viewers of the fused data set categorized as not watching the respective television advertisement of the plurality of television advertisements;
identifying specific characteristics of each media property of the identified media properties;
generating a targeted list of media properties based on the identified specific characteristics of the identified media properties;
generating an advertising campaign of de-duplicated reach comprising the targeted list of media properties;
executing the advertising campaign of de-duplicated reach by:
receiving an online ad impression from a remote server, the online ad impression comprising a media property from the targeted list of media properties and a first user; and
responding to the online ad impression by providing an advertisement to the remote server for display to the first user via the media property from the targeted list of media properties.
The application was rejected only under Section 112 (the examiner withdrew 101 rejections in the Answer). The examiner maintained that the specification did not provide adequate written description support for “generating an advertising campaign of de-duplicated reach comprising the targeted list of media properties.” According to the examiner, the specification “does not provide a disclosure of the computer and algorithm in sufficient detail to demonstrate to one of ordinary skill in the art that the inventor possessed the invention, including how to program the disclosed computer to perform the specifically claimed functions.” In particular, the Examiner found that the
specification only reiterates the claimed functions, but does not adequately disclose any particular algorithm, i.e., a series of process steps, to actually perform the claimed function.
The PTAB affirmed the rejection. Their analysis first confirmed that the disputed limitation broadly recited a function for achieving a desired result. The Board then went through each of Adobe’s arguments.
With regard to support in the specification, the Board found that the specification merely reiterated the same function. For example, Adobe pointed to disclosure of various factors considered to generate the campaign. In response, the Board found that:
Although the Specification describes what information is considered to generate the resultant advertising campaign, e.g., viewers who have or have not seen an advertisement, the websites visited by viewers, and associated linkages, merely disclosing the pieces used to calculate or determine a result does not explain what algorithm(s) or steps are taken in making that determination. In other words, the disclosure describes that certain identified information is considered, but no more than that.
It thus can be dangerous to have a tight relationship between the claims and the specification. An alternative drafting approach may be to have various levels of additional algorithmic details in the specification. Further, when amending, it may be advantageous to always keep the claim scope slightly broader than the most narrow algorithmic disclosure so that one can always point to a more detailed “algorithm” in the specification no matter how broadly the USPTO characterizes the alleged function. Of course, this is only possible if the specification has various layers and levels of details around the claimed feature.