0d352b4d3a317e3eae221199fdb49651-AuthorFeedback.pdf

We thank the reviewers for the valuable comments and discussions. Thissettingwas originallyproposedin the6 seminal work by Kempe et al. [19], and most follow-up studies adopt this particular setup (e.g. Therefore,12 our online influence maximization (OIM) is directly on the classical LT model, turning model parameters13 w(e) to be unknown (but fixed) and to be learned in an iterative manner. This is in parallel to the OIM for14 IC model [11,43,45], which also learns unknown edge probability parameters.15 Per ourabove16 discussion, we first want to clarify that the threshold on each node is not a model parameter of the classical17 LT model and our work is a frequentist approach for the onlinesetting.