| Summary: | With the development of the entertainment and film industry, people have more chances to access movies. Also, thanks to the population of online video websites, people prefer watching movies at home alone or with friends to going to the cinema. However, viewers may have different tastes. It is involuted to have some solid criterion on a ‘Good Film’. For a film recommendation app/website, the accuracy of recommending viewers what they like plays an important role.
Many film recommendation sites have their ranking systems which mainly based on the average users’ score. Some may tag different films like ‘horrible’, ‘comedy’, ’romantic’, and recommend films according to users’ viewing history. These two ways are common methods when recommending films.
In this thesis, we will focus on some recommendation methods based on machine learning. Factorization Machine, Attentional Factorization Machine, Wide & Deep Learning, and Deep Factorization Machine will be used in the dissertation and their advantages and disadvantages will be compared.
|
|---|