Skip to content

This project predicts community evolution in social networks

License

Notifications You must be signed in to change notification settings

AKheli/EPredictor

Repository files navigation

EPredictor

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.

Prediction Process

The process of predicting the evolution of communities consists of four steps:

  1. Segmentation of data: where the data is split into snapshots covering periods of time.
  2. Detection of communities: by extracting communities from each Snapshot.
  3. Identification of community changes: over time between each two consecutive snapshots and creation of evolution chains of communities.
  4. The prediction of future community evolution: by creating a prediction model created to predict the next event a community will manifest.

Motivation

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.

Contributing

Fork, implement, add tests, pull request, get our everlasting thanks and a respectable place here :).

Documentation

View the EPredictor User Guide.

Contact Us

Blog: http://lmcs.esi.dz

Mail: lmcs@esi.dz

License

EPredictor is released under the MIT License.