Recommender System ==================== This overview outlines the `Informfully Recommenders `_ repository. It can be used in combination with the Informfully Research Platform or in a stand-alone fashion. Informfully Recommenders is an extension of `Cornac `_. And you can look at the `Tutorial Notebook `_ for hands-on examples of everything outlined here. .. image:: uml/framework_extension_v4.2.png :width: 700 :alt: Overview of the extension stages The diagram above shows the extended framework Informfully Recommenders provides. It includes four dedicated stages: pre-processing, in-processing, post-processing, and evaluation. Additionally, it features a save state manager that allows the results of any stage to be stored and loaded. This allows for reusing existing data, e.g., applying different re-rankers to the same candidate list. Please find below an outline of the individual components, along with a link to their dedicated wiki pages for further information. Pre-processing Stage -------------------- * `Data Loading `_ * `Data Augmentation `_ In-processing Stage ------------------- * `Data Splitting `_ * `Participatory Diversity `_ * `Deliberative Diversity `_ * `Random Walks `_ * `Neural Baselines `_ Post-processing Stage --------------------- * `Re-rankers `_ * `User Simulator `_ Evaluation Stage ---------------- * `Metrics `_ * `Visualization `_