An Informative PTAB Decision on Patent Eligibility under 35 U.S.C. § 101
The U.S. Patent Trial and Appeal Board (“PTAB”) recently designated its decision in Ex Parte HANNUN (Appeal 2018-003323) (“HANNUN”) as being informative regarding the application of the latest 2019 revised guidance on patent-eligible subject matter. As a side note, the initial publication of the decision was erroneously titled Ex Parte Linden and included the wrong application number on the cover sheet, but the PTAB has since corrected this mistake. HANNUN is the seventh PTAB decision on statutory subject matter that has been designated as informative or precedential. Read more about prior PTAB decisions here.
HANNUN involves U.S. Application No. 14/735,002, now issued as U.S. Patent No. 10,540,957, which is directed to speech recognition systems developed using end-to-end deep learning. The technology is designed to improve speech to text transcription in noisy environments using a simplified and scalable architecture. At the time of filing, the application was assigned to Baidu USA LLC, which is the American R&D center for China’s largest search engine provider.
After the Examiner twice rejected all claims as being unpatentable under 35 U.S.C. § 101, and provided very little reasoning, Applicant appealed to the PTAB. In reversing the Examiner, the PTAB applied the Supreme Court’s two-step framework for analyzing patent-eligible subject matter (Alice test). Alice Corp. v. CLS Bank Int’l, 573 U.S. 208, 216 (2014). The analysis under the Alice test first determines whether a claim directed to an abstract idea, and if so, proceeds to the second step of determining whether there is an inventive concept sufficient to transform the abstract idea into patent-eligible subject matter.
In HANNUN, the PTAB noted that mental processes and methods of organizing human activity are two separate categories of abstract ideas. The claims recited steps of normalizing an input file, generating a jitter set of audio files, generating a set of spectrogram frames, obtaining predicted character probabilities from a trained neural network, and decoding a transcription of the input audio using the predicted character probability outputs. The PTAB concluded that these steps cannot be practically performed in the mind and are therefore not a mental process. The PTAB also found that these steps do not recite any methods of organizing human activity. The PTAB noted that while the claims may be based on mathematical concepts, the claims themselves do not recite any mathematical algorithms or formulas. As such, the PTAB determined the claims were not directed to any type of abstract idea.
Since the claims were not directed to an abstract idea, the analysis under the Alice test should have concluded, however, the PTAB took the opportunity to comment on the Examiner’s rejection. The PTAB noted that the Examiner was remiss for merely concluding that the claims fail to meet the second step of the Alice test without providing any factual support. The PTAB also reversed the Examiner’s determination of obviousness in view of the cited prior art, handing a complete victory to Applicant.
HANNUN provides another example of what types of activities should not be regarded as abstract ideas. For arguing against rejections alleging that the claims recite mental processes, it may be useful to point out the practicality of mentally performing the steps. For arguing against rejections alleging methods of organizing human activity, it may be useful to point out a lack of recognized, organizing human activities such as fundamental economic principles or practices, commercial or legal interactions, managing personal behavior or relationships or interactions between people. HANNUN seems to limit abstract idea rejections alleging mathematical concepts to instances where a mathematical algorithm or formula is affirmatively recited in the claims. Finally, HANNUN provides a reminder that Examiners need to provide factual support for rejections, and if they fail to, there is an option to attack the rejection itself.
Author
