Multi-Objective Blended Ensemble for Highly Imbalanced Sequence Aware Tweet Engagement Prediction
Published in 2020 ACM Recommender Systems Challenge Workshop, RecSysChallenge 2020, held at 14th ACM Conference on Recommender Systems, ACM RecSys 2020, 2020
Recommended citation: Felicioni N., Donati A., Conterio L., Bartoccioni L., Hu D.Y.X., Bernardis C., Ferrari Dacrema M. (2020). "Multi-Objective Blended Ensemble for Highly Imbalanced Sequence Aware Tweet Engagement Prediction." 2020 ACM Recommender Systems Challenge Workshop, RecSysChallenge 2020, held at 14th ACM Conference on Recommender Systems, ACM RecSys 2020. https://www.scopus.com/inward/record.uri?eid=2-s2.0-85092744154&doi=10.1145%2f3415959.3415998&partnerID=40&md5=3269bb23c4590d2992deb922beb21eed
Abstract: In this paper we provide a description of the methods we used as team BanaNeverAlone for the ACM RecSys Challenge 2020, organized by Twitter. The challenge addresses the problem of user engagement prediction: the goal is to predict the probability of a user engagement (Like, Reply, Retweet or Retweet with comment), based on a series of past interactions on the Twitter platform. Our proposed solution relies on several features that we extracted from the original dataset, as well as on consolidated models, such as gradient boosting for decision trees and neural networks. The ensemble model, built using blending, and a multi-objective optimization allowed our team to rank in position 4. © 2020 ACM. keywords: Blending; Decision trees; Multiobjective optimization; Recommender systems; Social networking (online); Ensemble modeling; Gradient boosting; Multi objective; User engagement; Forecasting pages: 29-33