A formal automated estimate of the index should be assessed and commented by a human expert. Similarity of 10-15% may be quite usual, especially due to reference lists, commonly referred papers. Extensive bibliographies produce high similarity.

Thank you

Result of tradition, then risk analysis and finally skin in the game: https://www.researchgate.net/publication/380427514_Kalergi_and_Hart-Cellerand_Memetics_White_Antifragility