An Anime recommendation system built using Unsupervised Learning (Nearest Neighbors) which recommends 5 closest anime to an anime the user likes
This is made using a Dataset from kaggle
Link of dataset : https://www.kaggle.com/CooperUnion/anime-recommendations-database
This dataset is made from myanimelist.net which is a website which contains an anime and manga database and you can provide ratings keep track and view what anime you could watch next.
It recommends the user 5 animes based on an anime the user likes.
It also recommends hentais.
The features considered while creating this model are :
- Genre (Slice of Life / Sport / Action / Drama etc. )
- Type (OVA / TV / Movie etc.)
- Members (The number of members that are there on myanimelist.net of the particular show)
- Rating (Rating is from myanimelist.net which has a public rated system and the ratings there are exceptionally accurate)
- Episodes ( number of episodes in the show)
The reason why episodes are considered while creating this model is because of the fact the many people like big animes (around 500 eps ) some like around (50 eps) and some like really short ones (12 eps). So episodes is also a major factor that needs to be considered.