You need a «.graphml» file to proceed further. In order to create it, please do not hesitate to read the brief introduction to NodeXL before starting applying GEPHI.
1. Download and install Gephi: http://gephi.github.io/users/download/
Open the graphml file saved in advance with the help of NodeXL.
2. Choose the layout algorithm - Force Atlas 2. Remember to prevent overlap and then hit Run. Pay attention to the form of the graph that has changed. Go to visualization. Use wheel to scale the
3. Use the statistics analysis (metrics). In the right field find Statistics and proceed with the measures.
4. As an example, the nodes differ in weight from each other according to the number of their friends. Betweenness centrality is a quantitative measure that calculates how many times a node Is situated on many shortest paths to go from one node to another. After you run the metrics, close the report. After this procedure, you may find the column of the metrics added to the datalab. Moreover, you may use it to visualize if you clique on range characteristics in accordance with betweenesscentrality. To draw a line, this would help you to find the central communication actors.
5. It appears to be quite helpful to calculate Modularity class metrics. After calculation go to Partition field. Refresh the criteria their and choose Modularity class. This measure helps to visualize the clusters by different colors.
6. Go to Datalab menu and copy «Tooltip» to «Label» column.
7. Then return to the tab of visualization and hit the T – text button.
8. *Remember to collect all nodes as few of them might be missed while using ForceAtlas 2 layout algorithm. Use filters of weight in the topology field for that aim. Filter the nodes that have the least weight. Then use the random layout and visualize those nodes with Force Atlas 2. Remember to remove the filter after that procedure.
9. Go to the Preview menu Refresh the graph. Export the file to PNG -preferences – 1024х1024
This guideline includes the procedure of using the basic functions of Gephi. The comparative analysis of social structure, social inequality and social communication within illustrated graphs is useful in detecting the similarities and differences between a number of social networks, the particular features in node connections and communication. In order to develop further it might be interesting to import the data from the facebook groups and analyze it with the same procedure.