Application of Data Mining to Predict the Likelihood of Contraceptive Method Use among Women Aged 15-49 Case of 2005 Demographic Health Survey Data Collected by Central Statistics Agency, Addis Ababa, Ethiopia
- *Corresponding Author:
- Tesfahun Hailemariam
Department of Health Informatics
Hawassa Health Science College
E-mail: [email protected]
Received Date: July 09, 2017; Accepted Date: July 13, 2017; Published Date: June 15, 2017
Citation: Hailemariam T, Gebregiorgis A, Meshesha M, Mekonnen W (2017) Application of Data Mining to Predict the Likelihood of Contraceptive Method Use among Women Aged 15-49 Case of 2005 Demographic Health Survey Data Collected by Central Statistics Agency, Addis Ababa, Ethiopia. J Health Med Informat 8:274. doi: 10.4172/2157-7420.1000274
Copyright: © 2017 Hailemariam T, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Introduction: In Ethiopia a gap between knowledge and use of contraceptive method is observed from many studies. According to the 2005 Ethiopian Demographic Health survey report the knowledge about any modern method among women is 86%, Contraceptive Acceptance Rate is 50.1% whereas the Contraceptive Prevalence Rate is 13.9%.
Methods: In order to find and interpret patterns from data the KDD process model is employed. This has gone through the steps of the process model; data selection and understanding, pre-processing, transformation, data mining, interpretation and evaluation. Decision tree and Naïve Bayes are used for the purpose of classification. The dataset used in this study is the 2005 demographic health survey data collected by central statistics agency. The techniques are tested both on the balanced and unbalanced datasets.
Results: Experimental results show that J48 decision tree performs better than Naïve Bayes. From this model 253 rules are generated. Overall accuracy of 82.85% a true positive (classifying non-user of contraceptive method) 87.3% and a true negative (classification of contraceptive method user) of 74.7% and a precision of 86.3%. One important rule detected was; women who do not know any contraceptive method have no any chance of using contraceptive method. But having knowledge of contraceptive method could not be a guarantee in order to use contraception. Other factors such as Partner occupation, Current marital status, wealth index, type of place were found to be most determinant factors as well.
Conclusion: Data mining techniques have revealed an important socioeconomic, demographic, geographic, reproductive history and knowledge factors associated with contraceptive method use. All concerned parties to strengthen the promotion of contraceptive method knowledge.