T 2049/12 - Functional data is not always technical

Key points

• This is an examination appeal concerning a patent application directed to software. 
• The Board: " A common misconception regarding the PHILPS decision [T 1194/97] is that there are only two kinds of data - cognitive and functional - and that functional (i.e. non-cognitive) data is always technical. The relevant question for assessing whether a data structure has technical character is rather whether it produces a technical effect. In the present case, the Board considers that there is no technical effect. Therefore, the claimed data structure does not provide an inventive step (Article 56 EPC)." 
• " Furthermore, the activity of programming is also excluded under Article 52(2)(c) and (3) EPC, because it is a mental act []. The choice of program structure, including the choice of data structures, belongs to the activity of programming. Well structured code helps the programmer in performing this activity, because the code is easier to understand, maintain, and adjust, but, since programming is not technical, this is not a technical effect." 

T 2049/12 (Data structure for defining transformations / MICROSOFT) - link

Reasons for the Decision

5. Inventive step (Article 56 EPC)

5.1 In the oral proceedings, inventive step was discussed starting from the prior art described in the application, from paragraphs [001] to [004]. As mentioned above, it was known to provide multiple services in a sequence, and to transform data provided from one service to another. In the prior art, those transformations (i.e. from "latitude" to "lat" and from "longitude" to "long") were implemented using "specific dedicated code" for each of the transformation. The dedicated code was, of course, stored on some computer-readable medium.

5.2 The prior art and the invention achieve the same transformations. The difference lies in how those transformations are defined. In the prior art, they are defined in "dedicated code", whereas in the invention, they are defined in a data structure having service identification fields, a transformation class field and transformation parameter fields.

5.3 The appellant argued that the claimed data structure provided more flexibility in defining the transformations. Owing to defining the transformations "declaratively", it sufficed to change the declaration in the data structure in order to change them; it was not necessary to change the code that carried out the transformation. The increased flexibility was a technical effect that supported the presence of an inventive step.

5.4 The Board, however, is not persuaded that the claimed data structure has technical character.

The invention in claim 1 provides a different, and arguably more flexible, way of structuring a computer program. Computer programs as such are excluded matter under Article 52(2)(c) and (3) EPC. Therefore, pure software concepts are considered to lack technical character (see for example T 1755/10 - Software structure/TRILOGY). Indeed, the software concept in TRILOGY was the separation of rules from an "engine" (the code that implements those rules), which is similar to the data structure and the transformations in claim 1.

5.5 Furthermore, the activity of programming is also excluded under Article 52(2)(c) and (3) EPC, because it is a mental act (see: G3/08, Reasons, point 13; and T 1539/09, Reasons, point 4.2). The choice of program structure, including the choice of data structures, belongs to the activity of programming. Well structured code helps the programmer in performing this activity, because the code is easier to understand, maintain, and adjust, but, since programming is not technical, this is not a technical effect.

5.6 The appellant furthermore argued that, when assessing a data structure according to established case law (see for example T 1194/97 - Data structure product/PHILIPS), a distinction had to be made between cognitive data and functional data. While cognitive data was non-technical, functional data had technical character. The claimed data structure was clearly functional data, and, therefore, it was technical.

5.7 The Board is not persuaded. The PHILIPS decision indeed makes a distinction between cognitive information content and functional data inherently comprising features of the technical system in which it is operative. However, in order to understand the distinction, one needs to look at the whole case.

The invention in PHILIPS concerned a picture data structure recorded on a record carrier. The data structure comprised information specifying where a relevant coded picture line had been recorded on the track, thereby facilitating access by a reader to the relevant picture content. The Board distinguished between the content of the picture, and the data structure, which had a technical function (facilitating access) in a technical system (reader plus record carrier). In other words, the data structure in PHILIPS mapped to technical features in a technical system.

In the present case, however, the data structure does not map to a technical system; it maps to a computer program, which is excluded matter under Article 52(2)(c) and (3) EPC.

5.8 A common misconception regarding the PHILPS decision [T 1194/97] is that there are only two kinds of data - cognitive and functional - and that functional (i.e. non-cognitive) data is always technical. The relevant question for assessing whether a data structure has technical character is rather whether it produces a technical effect. In the present case, the Board considers that there is no technical effect. Therefore, the claimed data structure does not provide an inventive step (Article 56 EPC).

5.9 In conclusion, the Board judges that the subject-matter of claim 1 lacks an inventive step (Article 56 EPC).

Order

For these reasons it is decided that:

The appeal is dismissed.