Graph Based and Content Based Filtering for Recommender Systems
| Project Type: | Research Project |
| Project Grant: | PITTA Grant 2017 (International Indexed Publication For UI Student's Final Project Grant) |
| Years: | 2017 |
| Organization: | Universitas Indonesia |
Description
There were three studies conducted under this project. The first study aims to determine the extent to which the combination of user similarity and item similarity is able to improve the performance of the recommendation system. This study explored several similarity formulations such as Jaccard Coefficient, Adamic Adar, Common Neighbor and Preferential Attachment and see which combinations can improve the performance of the recommendation system. The second study aims to see how attribute extraction for content-based approach can improve the performance of recommendation systems. The third study aims to use graph to detect fake news source in online social network.
Roles and Responsibilities
- Roles: Research Supervisor, Research Member
- Responsibilities: My responsibilities in this project includes: developing research design, supervising student in conducting research, composing paper/article for publication, and writing progress and financial report