College of Natural and Computational Sciences School of Information Science
| dc.contributor.advisor | Yifiru, Martha (PhD) | |
| dc.contributor.author | Getaneh, Desalegn | |
| dc.date.accessioned | 2019-05-24T11:52:19Z | |
| dc.date.accessioned | 2023-11-18T12:44:23Z | |
| dc.date.available | 2019-05-24T11:52:19Z | |
| dc.date.available | 2023-11-18T12:44:23Z | |
| dc.date.issued | 2017-10-04 | |
| dc.description.abstract | Ethiopia is one of the most populous country in Africa with a fertility rate of 4.6 and eight percent unwanted pregnancy. This contributes for maternal mortality and child death. To make every child is wanted, family planning plays an essential role through delaying, spacing or limiting birth. But the contraceptive prevalence rate of Ethiopia is low and needs to work hard to address the demand of family planning through providing choice of contraceptive method to a woman. Socio demographic characteristic has a contributing factor for choice of contraceptive method but which variables determine the choice of contraceptive method is a challenge. Moreover, among clients received contraceptive method, they might not get their choices. Due to this, there is 13 percent discontinuation rate though service provider uses different job aids to support the choice of contraceptive method. Thus, this research focus on identifying key variables that determine the choice of contraceptive method though applying data mining techniques and develop a knowledge based system that supports the health service provider for the choice of contraceptive methods. Empirical research design is applied to achieve this objective which combine both experimental and non-experimental researches. Prototyping approach is followed to develop the knowledge based system. As a research method, knowledge engineering and hybrid data mining methodology was employed. Interview and document analysis was also conducted to acquire knowledge from domain experts and documents respectively. A decision tree J48 algorithm was used to predict variables that determine the choice of contraceptive method. Thus, client age, number of children, education, residence, marital status, religion, region and contraceptive history determine the choice of contraceptive methods are key variables in the choice of contraceptive method. In addition, medical eligibility criteria and life style of a woman has a factor in the choice of contraceptive method. Prototype knowledge based system is developed that determine choice of contraceptive methods through integrating data mining results (socio demographic variables), medical eligibility criteria (explicit knowledge) and life style of a woman (tacit knowledge). x Based on system performance evaluation and user acceptance test, 86.6 % of accuracy and 76% acceptance was scored respectively so that integrating socio demographic data, medical eligibility criteria and life style of a woman is possible and can be implemented in the domain area. Finally, further exploration has to be done to refine the knowledge base and boost the advantages of choice of contraceptive method of knowledge based system to incorporate woman who has special characteristics. | en_US |
| dc.identifier.uri | http://etd.aau.edu.et/handle/12345678/18317 | |
| dc.language.iso | en | en_US |
| dc.publisher | Addis Ababa University | en_US |
| dc.subject | Family Planning | en_US |
| dc.subject | Choice of Contraceptive Method | en_US |
| dc.subject | Data Mining | en_US |
| dc.subject | Hybrid Modeling | en_US |
| dc.subject | Knowledge Based System | en_US |
| dc.subject | Contraceptive Method Choice Knowledge Based System | en_US |
| dc.title | College of Natural and Computational Sciences School of Information Science | en_US |
| dc.type | Thesis | en_US |