/friend-suggest-algorithm

Third party implementation of Suggesting Friends using the Implicit Social Graph

Primary LanguagePythonMIT LicenseMIT

Friend-Suggest-Algorithm

Third party implementation of the paper Suggesting Friends using the Implicit Social Graph for a user egocentric network.

Usage

python fsa.py

Sample Data (for Evaluation)

A custom/internal email dataset was used in the original paper to evaluate the algorithm. Since the dataset is not public, the Enron email dataset was used here. The Enron dataset is a standard public email dataset. You can download the dataset from here. The script data_loader.py was specifically implemented to load the some emails (outgoing/incoming) of a single user account from dataset.

Steps to use the (Enron) Data

  • Unzip the downloaded dataset
  • From a single user folder, select these four folder: inbox, notes_inbox, sent, sent_items
  • Copy these four folder to into the dataset folder

Requirements

  • Python3

Note

The code was written with verbosity (in mind) over speed.

License

  • All code in this repository are licensed under the MIT license as specified by the LICENSE file.