Earticle

초록

영어

In spite of a vast amount of extant research on recommender systems, two important questions did not receive much attention: 1. Should the missing ratings be estimated based on the mechanism of missingness ? and 2. Are the existing findings appropriate for businesses with relatively medium-sized inventories of up to 20,000 items ? Our study addresses these questions by experimenting with ten publicly available datasets. This unique 3 x 3 x 5 study explores three missingness mechanisms, three levels of missingness and five methods. With an empirics-first methodology, the study reveals the limitations of two fairly popular methods. It also offers a few striking results which should be of interest to businesses with medium-sized inventories.

목차

ABSTRACT
Ⅰ. Introduction
Ⅱ. Literature Review
2.1. Ratings Estimation
2.2. Mechanisms of Missingness
2.3. Research Gaps
2.4. Research Questions
Ⅲ. Research Methodology
3.1. Data
3.2. Methods
3.3. Evaluation Metrics
Ⅳ. Result Analysis
4.1. Results of MCAR Missingness
4.2. Results of MAR Missingness
4.3. Results of MNAR Missingness
4.4. Robustness Check
4.5. Recommended Approaches
Ⅴ. Discussion
5.1. Methodological Implications
5.2. Managerial Implications
Ⅵ. Limitations and Future Directions
Ⅶ. Conclusion
Disclosure Statement

키워드

저자

• Anand Va [ Assistant Professor, BITS Pilani, India ] Corresponding Author

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  Asia Pacific Journal of Information Systems

• 간기

  계간

• pISSN

  2288-5404

• eISSN

  2288-6818

• 수록기간

  1990~2026

• 등재여부

  KCI 등재,SCOPUS

• 십진분류

  KDC 325 DDC 658