| dc.description.abstract | The influx of asylum seekers and refugees, especially in recent times, has caused a bottleneck in Utlendingsdirektoratet (UDI)'s application process, leading to long waiting periods for their work and residency permits. During this time refugees have difficulties finding pastime activities, and instead of having possibilities of integration and inclusion, they suffer from boredom, passivization and reduced quality of life. This paper researches possibilities for using techniques from recommender systems to aid in solving this problem of finding activities and integrating refugees into Norwegian society. After researching state of the art tools and techniques and current existing systems pertaining to similar areas, a definite trade-off between simplicity of the system, and the accuracy and effectiveness of its recommendations was observed in existing solutions. This resulted in a content-based recommender system, Refugenious, being developed and tested, where the idea is to recommend activities without requiring vast amounts of user data. The initial results of its evaluation are promising and show that such a recommender system can provide good recommendations with only a small amount of information. The recommender system also proves to aid with lowering the threshold for finding activities, increase participation and widen refugees' specter of interests. Refugenious is planned to be deployed on the web portal Jibe within the end of 2017. | |
