Recommendation system in selecting course of public university in Malaysia using K-nearest neighbour
Internet began growing up with immense speed in these recent years. Business field had discovered many customers and income in the internet. There are huge information and items, and it became overload. Recommendation systems is developed to overcome this problem. Nowadays, freshly school leavers es...
| Main Author: | |
|---|---|
| Format: | Undergraduates Project Papers |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/26825/ http://umpir.ump.edu.my/id/eprint/26825/ http://umpir.ump.edu.my/id/eprint/26825/1/Recommendation%20system%20in%20selecting%20course%20of%20public%20university%20in%20Malaysia%20using%20k-Nearest.pdf |
| Summary: | Internet began growing up with immense speed in these recent years. Business field had discovered many customers and income in the internet. There are huge information and items, and it became overload. Recommendation systems is developed to overcome this problem. Nowadays, freshly school leavers especially for STPM and Matriculation college students in Malaysia have trouble with selecting a suitable course of public university. There are 20 public universities offering 893 bachelor courses in Malaysia. They need to apply their interested courses through UPU online within stated deadline. In Malaysia, there is only iMASCU which is the recommendation system that show and check the qualification of courses based on the qualification of user. K-Nearest Neighbour based recommendation system is implemented to solve this problem. The objectives of this project are to study the current algorithm and technique in recommendation systems for selecting courses; to implement k-Nearest Neighbour in the recommendation system; and to evaluate the application of the recommendation system. The data is collected from survey made to current public university students. There are six inputs and one target output in the training and testing sets. The result showed that the desired k value for the data is one and k-Nearest Neighbour can be implemented in the Malaysia public university course selection. |
|---|