摘要：The objective of the research reported here is to ascertain what variables influence the behavior of the patents that receive the most citations. The literature recognizes the number of forward citations as one of the most important indicators for discovering an invention’s value. This claim is tested in a research population consisting of the set of Brazilian green patents registered in the Global Patent Index database. The following variables are used: applicant type, applicant country, number of inventors, International Patent Classification code, family size, renewals, Patent Cooperation Treaty applicability, triadic family, grant and number of citations. These variables are used to create a decision tree and a logistic regression with the Statistical Package for the Social Sciences, V.21, and MS Excel. The results show that, in the behavior of Brazilian green patents, family size and renewals are connected to the mostcited patents. Another significant variable, but one whose numbers were not significant enough to place it in the decision tree, was membership in the Patent Cooperation Treaty.
