Actually, the relationship of the rank matrix  \(F\) and \(S\) represented that of the similarity matrix  \(\mathbb{F}\) and \(\mathbb{S}\), and we calculated the approximation degree of  \(F\) and \(S\) to demonstrate the reliability of the property relation map. The specific calculation process is as follows: for each target task, we obtained the corresponding rank sequences from the rank matrix  \(F\) and \(S\), and calculated the cosine similarity of the two above sequences by equation (2):