20 January 2010

Patentability of an apparatus conducting a mathematic algorithm

Besides the Bilski case (see here and here), in which the Machine-or-Transformation-Test (MOT) is used to verify whether or not a method claim represents statutory subject-matter, the decision ex parte Gutta of the US Board of Patent Appeals (BPAI) recently determined criteria to verify the same for an apparatus claim.

The application particularly relates to a business method an apparatus for conducting the same, whereas the business method claims have already been rejected due to not meeting the Bilski standard (i.e. the MOT-Test). What is interesting about this decision is the fact that the apparatus claims have been rejected under § 35 USC 101 (non-statutory subject matter) based on the reasoning, that the apparatus may be reduced to the underlying mathematical algorithm which, as such, is not patentable (see here and here). This ruling will surely influence the USPTO's practise on examining apparatuses conduction a CII, e.g. computers or the like.

The ruling of the BPAI was base on a new test, according to which an apparatus is only patantable if at least one of the the two following questions can be positively be answered:

  1. Is the claim limited to a tangible practical application, in which the mathematical algorithm is applied, that results in a real-world use (e.g., not a mere field-of-use label having no significance)?
  2. Is the claim limited so as to not encompass substantially all practical applications of the mathematical algorithm either “in all fields” of use of the algorithm or even in “only one field?”
Claim 14 of the application, which is directed to a "system for identifying one or more mean items for a plurality of items" structually only mentioned a memory and a processor, the latter being configured to "compute a variance of the symbolic values of the plurality of items relative to each of the items" and to "select the at least one mean item having a sysmbolic value that minimizes the variance", that is, to execute a pure mathmatical algorithm.

The BPAI thus found that for this subject-matter each of the above two questions has to be answered negatively, since the claim is neither "limited to a tangible practical application in which the mathematical algorithm is applied that result in a real-world use”, nor is the claim limited such that it does "not encompass substantially all practical applications of the algorithm".

Noteworthy, claim 19, which is directed to an article of manufacture comprising a computer-readable program code (for identifying the mean item having the symbolic value minimizing the variance) embodied on a computer-readable medium, has been accepted since the "computer-readable medium" (RAM, ROM or the like) represents a tangible entity. Hence, ex parte Gutta (again) opens the way to protect pure software (or mathematical algorithms) as stored on a computer-readable medium, which vaguely reminds Europeans of the EPO Board of Appeal decision T 1173/97 "computer program product/IBM" from 1998.