Skip to content

Latest commit

 

History

History
2 lines (2 loc) · 558 Bytes

File metadata and controls

2 lines (2 loc) · 558 Bytes

Aspect-based Sentiment Analysis of Scientific Reviews - JCDL 2020

Our JCDL 2020 paper illustrates a variety of finds based on openreview corpus. Most importantly it shows that emperical/theoretical soundness, impact of ideas and clarity are the most important attributes reviewers look for in a paper to recommend acceptance or rejection. Our paper also shows that disagreement among reviewers happen the least deciding the appropriateness of the paper to the conference. On the other hand, disagreement is highest for categories like novelty/originality.