This project focuses on uncovering social structures and user behavior patterns within derStandard.at by applying advanced network analysis techniques.
Through the use of multiple community detection algorithms and efficient data processing pipelines, it enhances content targeting and engagement strategies.
- Led Community Detection Analysis: Conducted in-depth analysis of user interactions to identify key communities and social trends.
- Applied Multiple Algorithms: Utilized six different network analysis algorithms to detect and map user communities, uncovering important engagement and behavior patterns.
- Enhanced Insights: Improved understanding of active user groups, supporting better content targeting and engagement strategies.
- Automated Data Pipelines: Developed and automated data workflows using R, igraph, and dplyr to process large-scale datasets reliably and efficiently.
+---------------------------+ +---------------------------+ +---------------------------+ +---------------------------+ +---------------------------+
| User Interaction Dataset | --> | Data Cleaning (dplyr) | --> | Graph Creation (igraph) | --> | Community Detection | --> | Insights & Reports |
| | | | | | | (6 Network Algorithms) | | |
+---------------------------+ +---------------------------+ +---------------------------+ +---------------------------+ +---------------------------+- Here are some output plots based on small dataset, that show detected communities using the menttioned algorithms. For more information, you can have a look at the report and R Markdown document.
|
 |
 |
|
 |
 |
- For channel categories with 100- 500 nodes, the following table shows comparative scores by using the above mentioned six algorithms:
