neo4jでsandbox その13

#概要

neo4jでsandboxやってみた。
movielens、やってみた。
レコメンドしてみた。

#コサイン類似度で映画をお勧め。

MATCH (p1:User {id:21})-[x:RATING]->(m:Movie)<-[y:RATING]-(p2:User)
WITH COUNT(m) AS numbermovies, 
	SUM(x.rating * y.rating) AS xyDotProduct,
	SQRT(REDUCE(xDot = 0.0, a IN COLLECT(x.rating) | xDot + a ^ 2)) AS xLength,
	SQRT(REDUCE(yDot = 0.0, b IN COLLECT(y.rating) | yDot + b ^ 2)) AS yLength,
	p1, 
	p2 
WHERE numbermovies > 10
WITH p1, 
	p2, 
	xyDotProduct / (xLength * yLength) AS sim
ORDER BY sim DESC 
LIMIT 10
MATCH (p2)-[r:RATING]->(m:Movie) 
WHERE NOT EXISTS((p1)-[:RATING]->(m))
RETURN p1.name, 
	p2.name, 
	m.title, 
	SUM(sim * r.rating) AS score
ORDER BY score DESC 
LIMIT 5

#python実行結果

title   score
Shawshank Redemption, The (1994)  29.503255423721242
WALL揃E (2008)  26.663597071412074
Inglourious Basterds (2009)  25.151535146327763
Whiplash (2014)  17.84941121812618
Shutter Island (2010)  16.417668233642118

以上。