Today we review yet another PTAB Section 101 decision where the USPTO’s revised guidance, and the Berkheimer Memo, show how examiners often fail to make appropriate Alice rejections - even when the patent is on nonsense… actual nonsense.
Specifically, the applicant here is IBM and they are aiming to patent nonsense. (We are not saying that the invention is nonsense, and thus not patentable subject matter under Section 101). In all fairness, IBM is actually aiming to patent the identification of nonsense. The case is SN 15/152,906, and the invention "relates generally to an improved data processing apparatus and method and more specifically to mechanisms for identifying nonsense passages in a question answering system based on domain specific policy."
Claim 1 on appeal is listed below:
1. A method, in a data processing system comprising at least one processor and at least one memory, the at least one memory comprising instructions executed by the at least one processor to cause the at least one processor to implement a question answering cognitive system, the method comprising:
receiving, by the question answering cognitive system, an input question from a user; obtaining, by the question answering cognitive system, a set of evidence passages for forming candidate answers to the input question from a corpus;
annotating, by an annotator in a nonsense identification component within the question answering cognitive system, a given evidence passage in the set of evidence passages with linguistic features to form an annotated passage, wherein a domain-specific policy is associated with a domain of the corpus;
counting, by a metric counters component in the nonsense identification component, a number of instances of each type of linguistic feature in the annotated passage to form a set of feature counts; determining, by the metric counters component of the nonsense identification component, a value for a metric based on the set of feature counts, wherein the metric is specified in the domain-specific policy;
comparing, by a comparator component of the nonsense identification component, the value for the metric to a predetermined model threshold, wherein the threshold is specified in the domain-specific policy;
determining, by a filter component of the nonsense identification component, whether the given evidence passage is a nonsense passage based on a result of the comparison;
responsive to the filter component determining the given evidence passage is a nonsense passage, sending, by the filter component of the nonsense identification component, the given evidence passage to a semi-structured data pipeline of the question answering cognitive system and preventing the given evidence passage from proceeding in the natural language processing pipeline;
responsive to the filter component not determining that the given evidence passage is a nonsense passage, passing, by the filter component, the given evidence passage to the natural language processing pipeline;
generating, by a hypothesis generating component of the question answering cognitive system, a set of candidate answers to the input question based on results of the semi-structured data pipeline and the natural language processing pipeline;
ranking, by an answer ranking component of the question answering cognitive system, the set of candidate answers to form a ranked set of candidate answers; and
outputting, by the question answering cognitive system, the ranked set of candidate answers to the user.
All joking aside, for all those who say that Alice is very predictable and is doing a great job of weeding out clearly unpatentable subject matter, it seems unlikely those voices would have or could have predicted a reversal in this case, especially one where no new ground of rejection or new reasoning to sustain a Section 101 rejection was presented by the Board.
Looking to the actual merits of this case, there is quite a bit of detail that seems technically-driven and relates to real world practical applications. IBM argued details about how the semi-structured data pipeline and natural language processing pipeline relate to specific technical issues in the technological field of question answering systems. The examiner simply relied on unsupported assertions about what would be conventional and well understood in the industry. In the end, the PTAB concluded that:
Because the Examiner has not provided sufficient evidence required by Berkheimer (Berkheimer, 881 F.3d at 1369), the Examiner erred with respect to Step 2B of the Guidance, and we are constrained by the record to reverse the Examiner's rejection of claims 1-10 and 14--23 on procedural grounds.
One cannot fault the PTAB here - on the record they really were constrained to reverse the rejection.
A procedural reversal is still a reversal and the examiner must now obtain approval from the Group Director if they want to re-open prosecution and issue new rejections. As of the date of the writing of this post, no action had yet been taken by the examiner. Whether a proper Alice rejection would hold up is left for another day.
So, even if you are trying to patent the identification of nonsense, applicants like IBM believe Alice is not necessary a block to patentability.