Social Media Analysis in a Nutshell
Download as PPTX, PDF1 like87 views
This document provides an overview of social media analysis and network science techniques. It discusses how to collect and graph data from Facebook and Twitter, and how to analyze the graphs using tools like NodeXL, Gephi, and SPSS. It describes how to identify influential users in a discussion and understand user roles. It also covers calculating basic network metrics, performing content and sentiment analysis, and examining individual user networks and how they are situated within broader networks. Finally, it discusses network interventions like stimulating influential users or groups, and optimizing network structure.
Social Media Analysis in a Nutshell
- 1. Social Media Analysis* *in a nutshell Nasri Messarra, Ph.D. American University of Beirut Université Saint Joseph 1
- 2. Getting and graphing the data Getting the data Graphing/Analyzing the data Facebook Twitter Facebook Twitter NodeXL NodeXL NodeXL NodeXL Graph Api Twitter Api Gephi Gephi Outwit (scraping) SPSS SPSS Facebook Analytics Manual Manual © Nasri Messarra, 2017 2
- 3. How it looks © Nasri Messarra, 2017 3
- 4. Finding the “mayors” of a Twitter discussion in a given period • قانوناألنتخاب © Nasri Messarra, 2017 4
- 5. Understanding the roles © Nasri Messarra, 2017 5
- 6. Calculating metrics 0 1 2 3 4 5 6 7 8 األنتخاب قانونGraph Metrics Maximum Geodesic Distance (Diameter) 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 األنتخاب قانونGraph Metrics Graph Density Graph Metric Value Graph Type Directed Vertices 474 Unique Edges 406 Edges With Duplicates 851 Total Edges 1257 Self-Loops 811 Reciprocated Vertex Pair Ratio 0.04040404 Reciprocated Edge Ratio 0.077669903 Connected Components 221 Single-Vertex Connected Components 167 Maximum Vertices in a Connected Component 93 Maximum Edges in a Connected Component 278 Maximum Geodesic Distance (Diameter) 15 Average Geodesic Distance 5.808979 Graph Density 0.001378221 Modularity 0.319885 © Nasri Messarra, 2017 6
- 7. Content analysis © Nasri Messarra, 2017 7
- 8. Sentiment analysis © Nasri Messarra, 2017 8
- 9. User’s network © Nasri Messarra, 2017 9
- 10. User in search network © Nasri Messarra, 2017 10
- 11. Network Interventions © Nasri Messarra, 2017 11
- 12. Network Interventions Stimulation Individuals (Champions) Segmentation (Groups) Induction (peer- to-peer) Alteration (Engineering) Adding/Deleting Vertices Adding/Deleting Edges Rewiring © Nasri Messarra, 2017 12
- 13. Misc. – Analysis of the network of two friends on Facebook 200 engaged fans as friends from stereotype 100 engaged fans as friends from “customer” Only 30 mutual friends © Nasri Messarra, 2017 13
- 14. Misc. – Comparison of the same discussion on two channels © Nasri Messarra, 2017 14
- 15. Twitter discussion about the garbage crisis Sep. 8, 2015 Civil society groups Community Clusters Twitter discussion about the garbage crisis Feb. 2, 2016 Two major groups: IranianAffairs and KSA24 Polarized crowds FromSmartMobstoSmartGovs? © Nasri Messarra, 2017 15
- 16. Optimization of networks 211 0 178 214 232 253 © Nasri Messarra, 2017 16