EPredictor is a tool of community evolution prediction in social networks using classification algorithms.
EPredictor grants the execution of prediction steps while dealing with the link between them, by starting the processus from any step, by importing and exporting multiple kinds of data and by visualizing results of different methods.
The process of predicting the evolution of communities consists of four steps:
- Segmentation of data: where the data is split into snapshots covering periods of time.
- Detection of communities: by extracting communities from each Snapshot.
- Identification of community changes: over time between each two consecutive snapshots and creation of evolution chains of communities.
- The prediction of future community evolution: by creating a prediction model created to predict the next event a community will manifest.
In their researchs, researchers perform each phase of the prediction process separately and they link manually between the inputs and outputs of the methods. The objective of EPredictor is to predict the evolution of communities within social networks.
Fork, implement, add tests, pull request, get our everlasting thanks and a respectable place here :).
View the EPredictor User Guide.
Blog: http://lmcs.esi.dz
Mail: lmcs@esi.dz
EPredictor is released under the MIT License.