A systematic literature review on attribute independent assumption of Naive Bayes: research trend, datasets, methods and frameworks

2nd Nommensen International Conference on Technology and Engineering

Ilham, Ahmad^1 ; Khikmah, Laelatul^2 ; Qahslim, Akhmad^3 ; Indra Iswara, Ida Bagus Ary^4 ; Laumal, Folkes E.^5 ; Rahim, Robbi^6

Informatics Department, Universitas Muhammadiyah Semarang, Semarang, Indonesia^1

Akademi Statistika Muhammadiyah, Semarang, Indonesia^2

Department of Information System, Universitas Al Asyariah Mandar, Polewali Mandar, Indonesia^3

Informatics Engineering Department, STIMIK STIKOM Indonesia, Bali Denpasar, Indonesia^4

Departement of Electrical Engineering, Politeknik Negeri Kupang, Indonesia^5

Sekolah Tinggi Ilmu Manajemen Sukma, Medan, Indonesia^6

关键词: Attribute independence assumption;    Data classification;    Development activity;    Finding solutions;    Literature reviews;    Research trends;    Software defect prediction;    Systematic literature review;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/420/1/012086/pdf
DOI  :  10.1088/1757-899X/420/1/012086

来源: IOP

【 摘 要 】

Recent studies of attribute independent assumptions on Naïve Bayes (NB) typically generate data sets, methods and frameworks that enable researchers to focus on development activities in terms of finding solutions to attribute independent issues, thereby enhancing the quality of NB classification and better utilizing resources. Many data sets deal with the NB attribute independence issues and different frameworks, so the overall picture of the independent assumption of the current NB attribute is not yet complete. This literature review aims to identify and analyze the research trends, data sets, methods and frameworks used in the attribute independence assumption research on NB for data classification between 2010 and 2018. The results of this research identified three frameworks that are highly cited and therefore influential in the software defect prediction field. They are Langley and Sage Framework, Friedman et al. Framework, and Wu et al. Framework.