Ist ein mathematischer Algorithmus patentierbar? (T1954/08)

In der Entscheidung T 1954/08 (Marketing simulation/SAP) vom 6. März 2013 hatte sich die Kammer 3.5.01 mit der Frage zu beschäftigen, unter welchen Voraussetzungen ein mathematischer Algorithmus patentierbar ist.

Mathematische Methoden und andere nicht-technische Gegenstände sind “als solches” per Gesetz vom Patentschutz ausgeschlossen. Eine erfinderische Tätigkeit kann deshalb nur durch einen nicht-naheliegenden technischen Beitrag (siehe die COMVIK-Entscheidung) begründet werden.

Ein mathematischer Algorithmus kann dieses Kriterium erfüllen, wenn er für einen technischen Zweck eingesetzt wird. Anders im vorliegenden Fall, wo der Gewinn einer Marketingkampagne prognostiziert werden sollte, d.h. der vorgeschlagene Algorithmus diente lediglich einem kommerziellen Zweck.

Aus der Begründung

[5.] In the light of Article 52(1)(2)(3) EPC, an inventive step according to Article 56 EPC 1973 requires a non-obvious technical contribution (T 641/00-Two identities/COMVIK, Headnote 1, OJ EPO 2003, 352; T 1784/06-Classification method/COMPTEL).

The use of computers for automation purposes is technical but commonplace.

A mathematical algorithm may become a technical means, i.e. it may go beyond a mere mathematical contribution, if it serves a technical purpose (T 1227/05-Circuit simulation/INFINEON, points 3.1, 3.2, OJ EPO 2007, 574).

[6.] However, anticipating a maximum revenue or profit value (A2, page 2, paragraph 2) of a marketing campaign is a commercial rather than a technical purpose. Therefore, the iterative mathematical algorithm of claim 1 remains a mere mathematical contribution which does not enter into the examination for an inventive step.

[6.1] The appellant argues that the choice of algorithm is based on technical considerations as it takes account of technical (e.g. memory) limitations of computers and diverges from a human approach.

[6.2] In decision T 1227/05, point 3.2.5 (supra) it was held that (the sole) processing speed was not a suitable criterion for distinguishing between technical and non-technical method steps since it was always possible to conceive of a slower algorithm than the one claimed. Similarly, the sole amount of memory a computer-implemented algorithm requires is equally unsuitable for determining whether or not a method step contributes to the solution of a technical problem since it is always possible to imagine an algorithm demanding more memory. Furthermore, whether or not an algorithm is similar to what a human being would do may play a role for the examination for inventive step, but this examination presupposes that the technicality of the feature has been established.

[6.3] The appellant further argues that the choice of algorithm is not part of the requirements supplied to the skilled person by the business manager, and concludes that the choice of algorithm is tied to a particular manner of (computer-)implementation.

The Board agrees that the mathematical algorithm is not provided by the business manager who is only interested in an economic forecast on which he can base his decision for a marketing campaign.

However, the Board does not agree with the appellant’s conclusion that the algorithm is provided by the implementing programmer. In the absence of a technical overall effect and purpose, the algorithm is provided by a mathematician for mathematical and ultimately commercial purposes. Mathematical definitions do not become technical by defining commercial relationships. For example, response probabilities and response values of customers are based on past customer behaviour (A2, page 26, last paragraph; page 27).

[6.4] As to the implementation of the algorithm, no internal function of the computer requires a non-obvious consideration to track and reverse incremental changes in the form of reassignments.

[6.4.1] The random number mentioned in claim 1 solves no problem other than splitting a large list of customers into two partial lists, without achieving any external technical effect or implying any technical consideration of the internal functioning of a computer.

On the implementation level, random number generators are well-known. The application implicitly confirms that finding as it is silent on any technical detail of generating random numbers.

[6.4.2] A binary map of flags settable for each possible customer-activity assignment (A2, page 8, lines 20/21) does not diverge fundamentally from a human’s approach when testing a multiplicity of assignments: a human would obviously mark (i.e. flag) tested assignments so as to test other assignments next.

Nor does the mathematical or commercial meaning of the flagged information imply any non-obvious technical modification of general computer functions.

[7.] The step that “the optimising algorithm is terminated after a user-defined number of customer reassignments does not improve the most recent best goal value” is considered next.

As mentioned above, the innovative potential of the algorithmic scheme can be left aside since it does not serve any technical purpose and, thus, does not contribute to the technical character of the claimed method and cannot enter into the examination for an inventive step.

Said lack of a technical purpose is not altered by defining a mathematical criterion for terminating the algorithm.

[8.] The Board concludes that the method according to claim 1 of the third auxiliary request does not involve an inventive step over a general computerised method for processing data according to any existing mathematical algorithm and, thus, does not meet the requirements of Article 56 EPC 1973.

Link zur Entscheidung T 1954/08 (Marketing simulation/SAP).

Leave a Reply