23
Social Network Analysis Using Gephi Nilkanth Shet Shirodkar Mtech – I 14103

Social Network Analysis Using Gephi

Embed Size (px)

Citation preview

Page 1: Social Network Analysis Using Gephi

Social Network Analysis Using Gephi

Nilkanth Shet ShirodkarMtech – I

14103

Page 2: Social Network Analysis Using Gephi

Social Networks

• A social network is a social structure of people, related (directly or indirectly) to each other through a common relation or interest

• Social network analysis (SNA) is the study of social networks to understand their structure and behavior

Page 3: Social Network Analysis Using Gephi

Network or Graph ?

• Network often refers to real systems Web, Socialnetwork, Metabolic network

Language: Network, node, link • Graph is mathematical representation of a

network Web graph, Social graph (Facebook or twitter)

Language: Graph, vertex, edge

Page 4: Social Network Analysis Using Gephi

• Communication networks: Intrusion detection, fraud detection• Social networks: • Link prediction, friend recommendation • Social circle detection• Social recommendations • Identify the network of an individual• community detection

Page 5: Social Network Analysis Using Gephi

Graph types

• Undirected graph– Facebook friendships

• Directed graph– Twitter: follow and be followed

Page 6: Social Network Analysis Using Gephi

Gephi

• Gephi is an open source tool designed for the interactive exploration and visualization of networks

• Designed to facilitate the user’s exploratory process through real-time analysis and visualization

• Visualization module uses a 3D render engine • Highly scalable can handle over 20,000 nodes

Page 7: Social Network Analysis Using Gephi

Graph measures

• Degree– In-degree– Out-degree

• Graph structure measures– Clustering (global and local)– Network diameter

• Centrality Measures– Eigenvector centrality– PageRank

• Community measures – Modularity

Page 8: Social Network Analysis Using Gephi

Degree Centrality

• In-Degree : The number of edges entering the node.

– Size of Node (Mention)

• Out-Degree : The number of edges leaving the node.

– Twitter User Mention other

Page 9: Social Network Analysis Using Gephi

StatisticsDegree: Calculate the number of links has a node

Degree weighted: Calculates the average number of links can

have node.

( Degree Distribution, In-Degree Distribution, Out-Degree Distribution )

Diameter : is the longest distance between two network nodes.

Closeness centrality: Measures the average distance between a node and all other nodes.

Page 10: Social Network Analysis Using Gephi

• Density : determines the percentage of network complementarity.

• Modularity : identifying groupings to highlight the communities in a network

• Eigenvector centrality: measures the importance of a node in the network according to its connections

• Related Components: determines the number of connected components in the network

Page 11: Social Network Analysis Using Gephi

Gephi Processes

• 1. Open• 2. Layout• 3. Ranking• 4. Statistics• 5. Rank (Modularity, Indegree and Outdegree)• 6. Layout (Size Adjust)• 7. Labels• 8. Community Detection• 9. Filter• 10. Label Adjust• 11. Preview• 12. Export

Page 12: Social Network Analysis Using Gephi

Gephi: Layout

• From the Layout module on the left side, choose Force Atlas from the dropdown menu, then click Run

• Force Atlas makes the connected nodes attracted to each other and pushes the unconnected nodes apart to create clusters of connections

• Click Stop when it seems as if you have some distinct clusters of nodes

Page 13: Social Network Analysis Using Gephi

Force Atlas : Individual Nodes are Outside and Communities are coming to center

Page 14: Social Network Analysis Using Gephi

Gephi: Rank(Modularity)

• Ranking in the top left module, and click Choose a rank parameter from the drop- down such as Modularity, Indegree and out degree.

• Set Min Size to 10 and Max Size to 150 • Set Min Size to 5 and Max Size to 200 • Min and Max size depends on the nature of your

network.• Click Apply to change the node sizes according to

their Modularity

Page 15: Social Network Analysis Using Gephi

Modularity Ranking

Page 16: Social Network Analysis Using Gephi

Gephi: Community Detection

• Go back to the Statistics tab on the right and click Run next to Modularity

• This creates a modularity class value for each node, which we’ll use to colorize the communities

• Click Apply to colorize the detected communities

Page 17: Social Network Analysis Using Gephi

Gephi: Filter

• Go to Filters in the top right module and open the Library Partition Count Modularity Class

• Filter option basically removes the “leaves” in the network that are not connected to many other nodes

Page 18: Social Network Analysis Using Gephi

Gephi: Label Adjust

• The Gephi recommended to run a final layout adjustment before the export that makes it easier to read the labels.

• “Label Adjust” works much the same as the size adjustment, moving the nodes so the labels are readable

Page 19: Social Network Analysis Using Gephi

Visualized twitter network community clusters

Page 20: Social Network Analysis Using Gephi

Calculating Basic Network Metrics Basic network, node, and edge metrics can be calculated using the statistics window.

Page 21: Social Network Analysis Using Gephi

Laying out the Network in Gephi The network will load in a random cluster of nodes. The first step will be to choose a layout to make the network more visible. Choose layout option and select Force Atlas

Page 22: Social Network Analysis Using Gephi

• The Social Network Analysis is a useful and effective instrument for revealing the main specificity of the human's relationships of the social groups.

• Software Gephi is the applicable tool for visualizing revealed people's interactions and the relational dimension of the communities inside the social groups.

Page 23: Social Network Analysis Using Gephi

REFERENCES

• [1] M. Bastian, S. Heymann, and M. Jacomy. Gephi: An open source software for exploring and manipulating networks. In Proc. 3rd International Conference on Weblogs and Social Media (ICWSM09), pages 361-362. AAAI, 2009.

• [2] S. Bickel and T. Scheffer. Multi-view clustering. In Proc. 4th IEEE International Conference on Data Mining (ICDM 04), pages 19-26, 2004.

• [3] D. Cai, Z. Shao, X. He, X. Yan, and J. Han. Mining hidden community in heterogeneous social networks. In Proc. 3rd International Workshop on Link Discovery, pages 58-65, 2005.

• [4] C. Ding and X. He. K-nearest-neighbor consistency in data clustering: Incor- porating local information into global optimization. In Proc. ACM Symposium on Applied Computing (SAC04), pages 584-589, 2004.

• [5] D. Greene and P. Cunningham. Multi-view clustering for mining heterogeneous social network data. In Workshop on Information Retrieval over Social Networks, 31st European Conference on Information Retrieval (ECIR09), 2009.