meta learning-based frameworks adaptively balance auxiliary tasks (meta-path prediction) with the primary task (link

We thank all four reviewers for unanimous support for the paper and constructive comments. Overall, reviewers are positive about our contributions: [R5] "The proposed framework is very general and It is the'first' paper to do so.", The overall quality of this paper is good." Is the proposed method applicable to existing heterogeneous GNNs such as GTNs [1]? Y es, as [R4] pointed out, our framework can be applied to any GNNs in a plug-in manner.