Some journals send a similarity-check-reports before accepting any article for publication, this sometimes reach more than 30%. To what extent I need to decrease the similarity and what is the maximum level of similarity that can be accepted by the jouranls.