This repository contains several prototypes that use different similarity metrics and clustering techniques on different SPICE case studies.
-
Example 1: This example shows how to implement community detection based on the similarity between a property of users. In addition, this example shows how to apply a custom similarity measure to detect communities. This code detects up to 5 communities based on the emotions that users felt watching artworks (information saved in users_emotions.json).
-
Example 2: This example shows how to implement community detection based on the similarity between a property of users. In addition, this example shows how to apply a basic similarity measure to detect communities. Basic similarity metrics are detailed in class SimilarityCommunityDetection. This code detects up to 5 communities based on the emotions that users felt watching artworks (information saved in users_emotions.json).
-
Example 3: This example shows how to implement community detection using graphs to relate users. In this example, it is possible to apply Markov Clustering and Greedy Modularity algorithms. Data used in this example is emotions_graphs.json.