Outranking Relations based Multi-criteria Recommender System for Analysis of Health Risk using Multi-objective Feature Selection Approach

被引:3
作者
Kuanr, Madhusree [1 ]
Mohapatra, Puspanjali [1 ]
机构
[1] IIIT Bhubaneswar, Dept Comp Sci & Engn, Bhubaneswar 751003, Odisha, India
关键词
Recommender system; Multi-Criteria Decision Making; Feature selection; Multi-objective Genetic Algorithm; DECISION-MAKING; SENSITIVITY-ANALYSIS;
D O I
10.1016/j.datak.2023.102144
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The recommender system filters out important information from a large pool of information to set some important decisions in terms of recommendation. It has had an impact in almost every domain, including health, where it can extract information from massive amounts of digital data on patients to provide a better understanding of their clinical status and predict diseases. It is found that most of the real-world problems are dealing with multiple (conflicting) criteria, and handling those criteria in decision making is a challenging task too. In this context, this paper has proposed a health recommender system based on multi-criteria ranking and a multi-objective feature selection approach. Certain risk factors for the development of cervical cancer in women have been recommended by the proposed system, as well as some models for reliable prediction of this disease. The performance of the proposed system is tested and analysed by implementing it on the Cervical Risk Classification dataset with the three different ranking algorithms of Multi-Criteria Decision (MCDM) Making. It is found that the MOORA (Multi-objective Optimisation on the Basis of Ratio Analysis) MCDM algorithm outperforms the other two algorithms in terms of Precision@N, Recall@N, F1-score@N and Mean Reciprocal Rank (MRR)@N. The performance of the proposed system is also verified using four other benchmarking disease datasets, and statistical Anova and Wilcoxon tests have been done to validate the results.
引用
收藏
页数:23
相关论文
共 58 条
[31]   Medical recommender systems based on continuous-valued logic and multi-criteria decision operators, using interpretable neural networks [J].
Ochoa, Juan G. Diaz ;
Csiszar, Orsolya ;
Schimper, Thomas .
BMC MEDICAL INFORMATICS AND DECISION MAKING, 2021, 21 (01)
[32]  
Ojokoh B A., 2012, Int. J. Comput. Sci. Inf. Technol. Secur, V2, P1008
[33]  
Pacheco A.G., 2016, ANN DATA SCI, P1
[34]  
Pardalos P., 2001, EUROPEAN J OPER RES, V130, P233
[35]   Personalized Re-ranking for Recommendation [J].
Pei, Changhua ;
Zhang, Yi ;
Zhang, Yongfeng ;
Sun, Fei ;
Lin, Xiao ;
Sun, Hanxiao ;
Wu, Jian ;
Jiang, Peng ;
Ge, Junfeng ;
Ou, Wenwu ;
Pei, Dan .
RECSYS 2019: 13TH ACM CONFERENCE ON RECOMMENDER SYSTEMS, 2019, :3-11
[36]  
Pincay J, 2019, INT CONF EDEMOC EGOV, P47, DOI [10.1109/icedeg.2019.8734362, 10.1109/ICEDEG.2019.8734362]
[37]   A Binary Multi-Objective Chimp Optimizer With Dual Archive for Feature Selection in the Healthcare Domain [J].
Piri, Jayashree ;
Mohapatra, Puspanjali ;
Pradhan, Manas Ranjan ;
Acharya, Biswaranjan ;
Patra, Tapas Kumar .
IEEE ACCESS, 2022, 10 :1756-1774
[38]   An analytical study of modified multi-objective Harris Hawk Optimizer towards medical data feature selection [J].
Piri, Jayashree ;
Mohapatra, Puspanjali .
COMPUTERS IN BIOLOGY AND MEDICINE, 2021, 135
[39]  
Rana S., 2020, CONTENT BASED HLTH R, P215, DOI [10.1002/9781119711582.ch11, DOI 10.1002/9781119711582.CH11]
[40]  
Rathee S., 2019, P INT C ADVANCEMENTS