Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476406Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Zalinda Othman, Professor Dr. | |
| dc.contributor.author | Nur Suhailayani Suhaimi (P61042) | |
| dc.date.accessioned | 2023-10-06T09:17:53Z | - |
| dc.date.available | 2023-10-06T09:17:53Z | - |
| dc.date.issued | 2013-08-05 | |
| dc.identifier.other | ukmvital:85044 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476406 | - |
| dc.description | The examination question generator is a relatively new instrument in education for the purpose of constructing a set of questions automatically from a question bank. Generally, in an automatic examination question generator, questions are structured in cases within a case-bank. The Case-Based Reasoning (CBR) technique has been used widely to generate examination questions based on question sets. However, CBR has its limitations. For example, the retrieval accuracy or the retrieved solution might not be close enough to the desired solution. That can be caused by many problems such as a lack of cases stored in the case-bank and indexing problems. The objectives of this research are to propose a Genetic Algorithm (GA) approach as a CBR retrieval optimizer and to evaluate the retrieval performance between single and hybrid approaches. This research performed the four cycles of CBR phases which are retrieve, reuse, revise and retain, to achieve the objectives. GA was integrated into the retrieve phase to solve the retrieval problem and automatically increase cases in the case-bank. Optimized and high accuracy solutions can be retrieved and reused as the recommended solutions to the user. Higher accuracy solutions were measured in terms of similarity between the obtained solutions and the desired solutions. Then, the obtained solutions were revised in order to index each solution accordingly. Finally, the solutions were stored in the case-bank. The highest accuracy for the CBR approach was 77.5 per cent and the highest accuracy for CBR-GA was 92.5 per cent. The proposed CBR-GA approach obtained satisfactory results with higher accuracy compared to the single approach of CBR. This result shows the effectiveness of the GA optimizer in increasing the amount of indexed cases in the case-bank. The increased number of cases improved the performance of searching. This research has proven that incorporating GA into CBR increases the accuracy and diversity of candidate solutions.,Master / Sarjana | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Genetic algorithms | |
| dc.title | Hybridization of case-based reasoning and genetic algorithm in the retrieval of examination question | |
| dc.type | theses | |
| dc.format.pages | 132 | |
| dc.identifier.callno | Q338.8.N847 2013 | |
| dc.identifier.barcode | 002118 | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_85044+SOURCE1+SOURCE1.0.PDF Restricted Access | 2.18 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.