- Helping users find suitable rooms easily without specifying multiple search parameters
- Finding an available meeting rooms when users current favorite rooms are unavailable
- Recommending new rooms to users to be added in their favorite list
Develop a recommender model that can recommend rooms of interest to users.
- Room recommender service uses a content-based recommender model based on Aalto room booking data from ASIO system.
- Room parameters used: room_type, room_category, #persons, square_meters
- Model is a memory based model where each room is transformed into a numerical vector and room similarity is calculated by using cosine similarity (the angle between room vectors)
Technical documentation can be found in Aalto Git version control