Network Analysis
Co-Reviewer Network
A network has been constructed by linking movies that share common reviewers. The Louvain algorithm is implemented to create communities by optimizing modularity, allowing detection of densely connected groups of movies that share similar characteristics. To ensure the graph will have distinguishable communities, only movies with at least 3 common reviewers are included (edge weight ≥ 3). Mainstream movies are hard to separate due to their broad appeal and are thereby represented in the largest communities in the middle, movies included here are Star Wars, superhero movies and typical blockbuster movies. That said, many large communities seems to be thematically aligned. Some by genres such as horror/suspense, while others are connected through shared cultural interests, like the cluster of classic British literature adaptations, which contains movies like Jane Eyre and Pride and Prejudice.
Network Visualization
Community Ratings Distribution
Network degree distribution
Network Structure
- 10,897 nodes (movies)
- 25,988 edges
- Edge weight threshold: ≥ 3
Community Detection
- Louvain algorithm (python-louvain)
- Modularity: 0.767
- Nodes annotated with title, average rating, and price
