Skip to content
Abdelouahab Khelifati edited this page Jun 23, 2017 · 2 revisions

EPredictor

EPredictor is a project that consists of a tool that does the prediction of community evolution in social networks using classification algorithms.

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.

Objective

EPredictor allows the following functionalities: snapshots segmentation, communities detection, communities evolution identification over time and prediction of their future evolutions. Those steps are exectuted by dealing with the link between them, by starting the processus from any step, by importing and exporting all kinds of data and by visualizing results of different methods.

Contributing

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

Documentation

View the EPredictor User Guide.

Clone this wiki locally