Lets say, for example, my patent requires computing all elementary cycles in a graph. It happens to be known that all such cycles can be computed from a cycle basis. Keep in mind I am not trying to patent the algorithm for computing the cycle basis nor the algorithms that compute all cycles from the cycle basis. Rather let's say my patent involves representing a concrete application as a graph and that the method of the patent utilizes the known theorems about computing cycles. In this illustration, the theorem: "All graphs have a cycle basis and all elementary cycles are a linear combination of cycles" is known and proven. But what if I need to make use of some other theorem that is not proven. Must I prove it in the patent? And what if my proof has an error? Should I even discuss the correctness of the method by referring to a math theoretic result?
1 Answer
You do not need to prove that your method works. If you claim a method that doesn't work, it just means that your patent will protect something that's probably not worth very much (since it doesn't work).
Your specification should include a description of the steps that you might use (both aspects that you have invented and aspects that are already known), but it doesn't need to justify why you're using those steps or prove that they will work correctly.
In short, no, you do not need to discuss the correctness of the method.
answered Jan 31, 2016 at 1:57
-
Doesnt a patent have to be useful? And if i gave an incorrect method wouldn't that mean my patent is not useful– dan bCommented Jan 31, 2016 at 10:57
-
@danb, you are correct that, technically, an invention must be useful in order to be patentable. However, this "utility" requirement has taken on a very broad meaning, such that essentially anything can qualify. You will not have any trouble qualifying for a patent based on the utility requirement. Commented Jan 31, 2016 at 15:23