Two Stages Approach for Tweet Engagement Prediction

TL;DR

This paper presents a two-stage ensemble approach combining diverse feature extraction methods and XGBoost to predict tweet engagement, achieving a competitive ranking in the RecSys Challenge 2020.

Contribution

It introduces a novel two-stage framework integrating heterogeneous features and ensemble learning for tweet engagement prediction.

Findings

Achieved 22nd place in the RecSys Challenge 2020

Effectively combined handcrafted, knowledge graph, sentiment, and BERT features

Demonstrated the viability of staged feature learning and ensemble methods

Abstract

This paper describes the approach proposed by the D2KLab team for the 2020 RecSys Challenge on the task of predicting user engagement facing tweets. This approach relies on two distinct stages. First, relevant features are learned from the challenge dataset. These features are heterogeneous and are the results of different learning modules such as handcrafted features, knowledge graph embeddings, sentiment analysis features and BERT word embeddings. Second, these features are provided in input to an ensemble system based on XGBoost. This approach, only trained on a subset of the entire challenge dataset, ranked 22 in the final leaderboard.