How Google, Youtube make recommendations
Recommendations, classifications and predictions (Data Science, Data Mining, Text Mining, Machine Learning)
Youtube recommendations uses data about devices. Data about devices includes what videos I watch, what videos I searched arched, what terms I searched on Google.
Youtube then recommended videos based on what videos I watched, what videos I searched, what terms I searched on Google.
The videos are based on clicks and watches and satisfaction. If I searched sourdough bread, Youtube will recommend sourdough bread videos and the videos displayed are based on clicks and watches and satisfaction. Youtube will also show related videos like sourdough pizza.
If I searched multiseed sourdough bread, other multiseed sourdough recipes will also be recommended.
If I searched recipes, during chinese new year, a lot of people watched dumplings, youtube will recommend dumplings.
During pneumonia, a lot of people watch bread and sourdough, so sourdough and bread videos ate recommended.
Google news will also recommend news based on what videos I watched, what videos I searched, what terms I searched on Google.
Lazada also recommend products based on what videos I watched, what videos I searched, what terms I search on Google, what term I search on Lazada. Lazada showed proof banneton…
