The main purpose of this demo is to demonstrate how to train the vector representation of items using Word2vec and make item recommendations based on the similarity of item vectors. It mainly consists of 4 parts:
Prepare item sequences based on user behavior.
Train a CBOW model using the Word2Vec module of the gensim library.
Extract all embedding data and write it to chDB.
Perform queries on chDB based on cosine distance to find similar movies to the input movie.
A simple unittest for vector data insertion and querying.
The main purpose of this demo is to demonstrate how to train the vector representation of items using Word2vec and make item recommendations based on the similarity of item vectors. It mainly consists of 4 parts: