Review Report
MetaReview
The paper is well written. The originality of the paper resides in applying the AIK method to a new task, namely multi-label unknown intent detection. The obtained results are interesting. The paper is missing some information. The authors have provided them in their rebuttal.
Overall Recommendation
Review #1: 4/5; Review #2: 4/5
Postscript (by Yawen)
I need to improve my drawing skills…