by Artyom Rozhkov https://github.com/artjop/sna-cp

Description (CP)

Key Idea

• Try to analyze an text comments to find an object and evaluate their sentiment
• Created a bipartite graph with users, their objects, and their attitudes as nodes. The weight of each edge will indicate the tone of this attitude.
• Try to categorize users based on their unique attitude set.

Limitation

• Data Sampling Size: 20 only the first 20 posts • Geography: only global Moscow groups • Data Source: VK as most popular service only

Description of the research dataset

• Table with 13887 filtered rows and 9046 not-unique ids from 300 groups • Scrapped by VK API