Talk with your Neo4j DB

Here we shall try to do some cool stuffs with Neo4j

Dataset files Description

small_grouped_data.csv : Contains all the movies with their properties(columns). To load them in your database, use the Cypher Query below:

LOAD CSV WITH HEADERS FROM 'https://raw.githubusercontent.com/Foutse/Neo4j_experiences/main/dataset_movie/small_grouped_data.csv' AS l FIELDTERMINATOR ',' CREATE(m:MovieLens{M_movieId:toInteger(l.movieId), M_rating:apoc.convert.fromJsonList(l.rating),M_userId:apoc.convert.fromJsonList(l.userId), M_timestamp:l.timestamp,M_title:l.title,M_summary:l.summary,M_year:toInteger(l.movie_year),M_genres:apoc.convert.fromJsonList(l.genres), M_Avg_rating:toFloat(l.Avg_rating),M_unique_tags:apoc.convert.fromJsonList(l.unique_tags), M_unique_Uid_tags:apoc.convert.fromJsonList(l.unique_Uid_tags),M_round_ratings:apoc.convert.fromJsonList(l.round_ratings), M_rating_vector:apoc.convert.fromJsonList(l.rating_vector),M_round_ratingsup:apoc.convert.fromJsonList(l.round_ratingsup), M_round_ratingsdown:apoc.convert.fromJsonList(l.round_ratingsdown),M_rating_vectorup:apoc.convert.fromJsonList(l.rating_vectorup), M_rating_vectordown:apoc.convert.fromJsonList(l.rating_vectordown)}) RETURN count(m)

genre_rel_all.csv : contains the relationship between movies based on common genre. The weight represents the jaccard similarity of their genre list. You can load the relationship to your AuraDB as follows:

LOAD CSV WITH HEADERS FROM 'https://raw.githubusercontent.com/Foutse/Neo4j_experiences/main/dataset_movie/genre_rel_all.csv' AS l FIELDTERMINATOR ',' MATCH (m1:MovieLens {M_movieId: toInteger(l.source)}), (m2:MovieLens {M_movieId: toInteger(l.target)}) WHERE m1 <> m2 AND toFloat(l.weight) > 0 MERGE (m1)-[j:COMMON_GENRES]-(m2) ON CREATE SET j.weight = toFloat(l.weight) RETURN count(j) as total_relations The same logic applies for the other relationship files.

user_rel_all.csv : Relationship based on common users who watched the movies
usertag_rel_all.csv : Relationship based on common users who tagged the movies
summary_textSimBert_all.csv : Relationship based on similarity between the text summary of the movies, fixed to a certain treshold
rating_vectordown_rel_all.csv : Relationship based on similarity between their rating vector
movie_tag_rel_all.csv : Relationship based on the common tags they have