22 August 2017 – EPO Case Law – case T 1028/14 – Reputation-based method and system for determining a likelihood that a message is undesired: Identification of technical and non-technical features

Claim 1 of application EP07816059.5 reads as follows:
A method of determining a likelihood that a received message is an undesired message, comprising the steps of:
(i) receiving a message at a messaging system;
(ii) forwarding to a reputation engine a pre-selected set of identifiers relating to the origin of the message, the pre-selected set of identifiers including an IP address from which the received message originated, a tuple of a domain at which the received message allegedly originated and the IP address from which the received message originated, and a tuple of a user who allegedly originated the message and the IP address from which the received message originated;
(iii) checking databases at the reputation engine to determine previously determined reputation metrics for the forwarded identifiers and returning any previously determined reputation metrics to the messaging system;
(iv) making a first determination at the messaging system of a likelihood as to whether the received message is undesired using a first set of criteria including the returned reputation metrics; and
(v) marking the message as being either desired or undesired in accordance with the first determination.
The examining division refused the European patent application on the ground of lack of inventive step, considering features (i) to (iv) of claim 1, leaving aside the terms “messaging system”, “reputation engine” and “IP address”, represented non-technical features merely relating to “a scheme for processing data”.
The board did not agree.
The established jurisprudence of the Boards of Appeal considers that a non-technical feature of a claim is a feature relating to non-inventions within the meaning of Article 52(2) EPC (including schemes, rules and methods for performing mental acts or doing business, and programs for computers) and does not contribute to the solution of any technical problem by providing a technical effect, even when interacting with the technical features of the claim; it can thus be ignored for the purposes of assessing inventive step.
According to the application of this case, the problem to be solved by the invention is to provide a reputation-based system and method for determining a likelihood that a message is undesired which permits finer granularity in tracking reputations. Given that the whole application is concerned with telecommunication messages and that undesired messages correspond to spam messages, the board deemed that the underlying problem to be solved is indeed a technical problem and that features (i) to (v) of claim 1 do in fact contribute to the solution of that technical problem, for the following reasons:
By feature (i), an apparently technical unit, i.e. a “messaging system”, receives a message including identifiers such as IP (Internet Protocol) addresses.
Through feature (ii) the messaging system forwards those identifiers to a unit, named “reputation engine”.
By feature (iii), that unit checks its databases to determine some metrics relating to the “reputation” of the identifiers forwarded and sends them to the messaging system.
Through feature (iv), the messaging system calculates the likelihood that the received message is an undesired message (i.e. spam message) based inter alia on the metrics delivered by the reputation engine.
Lastly, by feature (v), the messaging system marks the message as desired or undesired.
Thus, by the use of those features, a more fine-grain assessment of whether or not a spam message has been received may indeed be achieved. Moreover, the method steps according to features (i) to (v) are indeed part of the functional computer system and do not merely relate to a “scheme for processing data” (which is in any event not mentioned in the list of non-inventions in Article 52(2)(a) to (d) EPC).
In conclusion, the board held that all the features (i) to (v) are technical features which consequently are to be considered in full in the assessment of inventive step. Therefore, the question is not whether a notorious networked computer system is able to perform the method steps of features (i) to (v) but whether the skilled person in the field of data processing would come up, without exercising inventive skills, with a solution enabling such a “networked computer system” to perform all those steps.
The board deemed that the examining division provided only an inadequate and flawed assessment of inventive step, citing only document D1 as evidence of the notoriety of a “generic networked computer system”. Therefore, the board did not consider itself in a position to assess the correctness of the examining division’s assessment of novelty and inventive step, or to pass final judgment on that issue for the very first time in the appeal proceedings. Accordingly, the board deemed that it could not accede to the appellant’s request for allowing the claims at this stage and decided to remit the case to the examination division for further prosecution, having regard to the pertinent prior art – whether cited in the application as originally filed or in the examination proceedings or to be identified following an additional search.
Finally, the board decided that the decision to remit the case to the department of first instance did not adversely affect the appellant, and could therefore be taken without having to appoint oral proceedings before the board. – Full decision