I have a paper that was a published in a conference proceedings. This paper was extended and turned into a journal paper. However, some things like figures from the motivation section or the experimental setup (used same metrics) are the same. The journal paper was reviewed, and one reviewer raised some concerns about a possible overlap with the conference paper. At the moment I'm preparing a response to his comments, and would like to present a percentage of the overlap. Is there a method to calculate the overlap with a previous publication? Are there any common guidelines that are used when presenting an overlap with a previous publication?
Your problem is that you need to show your paper is sufficiently original to merit publication. Calculating a measure of overlap will not help you achieve that goal. Instead, you need to:
- Clarify the original aspects of the paper. For example, you might write in the paper "Previously we showed [minor progress]" and then "but we did not solve [major problem in the field]" and finally "in this work we achieve [important step towards goal]"
- Justify to the editor why some overlap is necessary. For example, "This paper is about x, which is well known to experts but may be unfamiliar to some readers of this journal. Therefore, we repeat our introduction to x from our conference paper."
- Remove unnecessary overlap.
The editor will be looking for meaning, not quantity. Don't invent metrics where they are not needed.
You should be able to list the major points in both papers.
If the lists are mostly identical then the second paper is too similar to the first, if the points are substantially different then the second does not overlap but takes the topic forward.