원문정보
초록
영어
Human activities have been investigated and applied in various fields, such as context-aware computing, search engine, social network services, location-based services, automated visual surveillance, and multimodal human–computer interaction. However, the human activities in the review process of academic conferences have seldom been explored. There is no doubt that review process plays an important role in deciding whether a paper can be accepted or not. In this paper, we present our work to understand the review activities by analyzing the anonymized review data of two conferences (ACM SIGCOMM and UIC). The descriptive statistics and the data mining technology are adopted in the analysis. We got some interesting knowledge, which is significant for interpreting how the reviewers give their reviews in academic conferences, such as the relationships between the score, confidence and review length, and reviewer activity patterns.
목차
1. Introduction
2. Related Work
3. Data
3.1. Data Overview
3.2. Data Preprocessing
3.3. Correlation between Score, Confidence, and Character Count
4. Data Analysis
4.1. Descriptive Statistical Method
4.2. Data Mining Method
5. Results
5.1. Review Patterns
5.2. Reviewer Activity Patterns
6. Conclusion
Acknowledgements
References