Projects
Buka tautan projectKembali ke daftar
ProjectSocial Network Analysis2024—
Social Network Intelligence
Enterprise-grade social network analysis platform for monitoring individuals and organizations across social media. Collects public conversations, performs sentiment analysis, identifies entity relationships, and visualizes complex interaction networks through an interactive real-time graph interface.

Tech Stack
- Go
- Gin
- Elasticsearch
- MongoDB
- WebSocket
- Gephi
Key Features
- Interactive network graph visualization with dynamic node/edge rendering and real-time WebSocket updates
- Sentiment analysis classifying social media content as positive, neutral, or negative
- Community detection using Gephi for modularity analysis, cluster detection, and centrality calculation
- Real-time graph data streaming via WebSocket for seamless live visualization
Architecture
Social media → graph processing pipeline
- 1.Social media content continuously collected and indexed in Elasticsearch
- 2.Relationship extraction from mentions, replies, reposts, and shared engagements
- 3.Gephi Toolkit processes the graph for community detection and layout generation
- 4.Processed graphs stored in MongoDB and streamed to clients via WebSocket
- 5.Golang (Gin) exposes REST APIs and WebSocket endpoints for the web application
Data / Processing Flow
- 01Social media data collected and indexed in Elasticsearch
- 02Relationships extracted from interactions (mentions, replies, reposts)
- 03Graph constructed: nodes = accounts/entities, edges = interactions
- 04Gephi engine performs modularity analysis and community detection
- 05Processed graph stored in MongoDB and streamed via WebSocket to clients
Highlights & Metrics
- Real-time WebSocket delivery of large-scale network graphs
- Community detection via modularity and degree centrality
- Elasticsearch + MongoDB dual-store for fast retrieval and persistence
Use Cases
- Person monitoring — track activity, influence, and relationships of individuals
- Organization monitoring — analyze presence, engagement, and network interactions
- Influence analysis — identify key influencers and central actors
- Community mapping — discover clusters around a specific topic
- Issue propagation analysis — understand how narratives spread through networks
My Contributions
- Designed the end-to-end graph processing architecture.
- Developed REST APIs and WebSocket services using Golang and Gin.
- Integrated Elasticsearch as the primary social media data source.
- Implemented graph storage and retrieval using MongoDB.
- Integrated Gephi Toolkit for graph generation and community detection.
- Optimized graph processing and real-time delivery performance.
Technical Highlights
- Gephi Toolkit integration connects a Java-based graph engine with Golang services
- Elasticsearch + MongoDB dual-store: fast retrieval for queries, persistence for graphs
- Real-time WebSocket delivery enables live graph updates without page reload
- Community detection via modularity analysis identifies clusters and influential actors
http://localhost:3000/projects/media-monitoring