EMiGRe: Unveiling Why Your Recommendations are Not What You Expect

Download PDF
Open Webpage

Hervé-Madelein Attolou, Katerina Tzompanaki, Kostas Stefanidis, Dimitris Kotzinos

Published: 2024, Last Modified: 16 May 2026ICWE 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: This demonstration showcases EMiGRe, a system tailored for computing explanations for missing recommendations in a graph-based recommendation system. The users can interact with the system through our intuitive visualization interface to navigate the graph, select their missing recommendations, choose their preferred explanation mode (add or remove user actions), and finally consume the explanations in textual or graphical form. n. Throughout, they are guided through important steps of EMiGRe (The source code for this demonstration is available here https://git.cyu.fr/hattolou/emigre_icwe2024) and useful statistics that clarify the scenario.