Social Network Analytics

Analyze social network dynamics — trending content, user influence, viral propagation, and engagement metrics — in a unified platform. Materialized views with three refresh modes (PERIODIC for trending dashboards, INCREMENTAL for post counts on commit, and MANUAL for on-demand influence scores) pre-compute key analytics, graph traversal powers viral spread chain analysis and community overlap detection, time-series tracking captures engagement metrics as historical snapshots, and polyglot query support lets you use SQL for views and aggregations alongside OpenCypher for graph traversals.

Architecture Overview

Vertices

User, Post, Topic, Group

Edges

FOLLOWS, CREATED, LIKED, SHARED, TAGGED, MEMBER_OF

Documents

EngagementMetric (postId, likes, shares, comments, timestamp)

Users create and interact with posts, follow each other, and belong to groups. Engagement metrics are captured as time-series snapshots. Materialized views pre-compute trending scores, post counts, and influence rankings.

Key Queries

Trending Content Dashboard — Aggregate engagement into trending scores (via PERIODIC materialized view):

SELECT postId, SUM(likes + shares * 2 + comments * 3) AS trendingScore
FROM EngagementMetric
WHERE timestamp > date('2024-01-15', 'yyyy-MM-dd')
GROUP BY postId ORDER BY trendingScore DESC

Viral Spread Chain — Trace how content propagates through the network:

MATCH path = (origin:User)-[:CREATED]->(p:Post)<-[:SHARED]-(sharer:User)-[:FOLLOWS]->(reached:User)
WHERE origin.name = 'Alice'
RETURN origin.name, sharer.name, reached.name, p.title

Community Overlap — Find users active in multiple groups:

MATCH (u:User)-[:MEMBER_OF]->(g1:Group), (u)-[:MEMBER_OF]->(g2:Group)
WHERE g1 <> g2
RETURN u.name, g1.name, g2.name

Try It Yourself

git clone https://github.com/ArcadeData/arcadedb-usecases.git
cd arcadedb-usecases/social-network-analytics
docker compose up -d
./setup.sh
./queries/queries.sh

Full source: social-network-analytics on GitHub