# Choice-Aware User Engagement Modeling andOptimization on Social Media

arXiv: https://arxiv.org/abs/2104.00801 · HTML: https://mlsanity.com/paper/2104.00801/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 7 of its 20 references that are in the map.
- First posted on arXiv 1 Apr 2021.

## Builds on (most cited first, 7 of 7)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Empirical Evaluation of Rectified Activations in Convolutional Network](https://mlsanity.com/paper/1505.00853/) (2015)
- [Assortment optimisation under a general discrete choice model: A tight analysis of revenue-ordered assortments](https://mlsanity.com/paper/1606.01371/) (2016)
- [Privacy-Aware Recommender Systems Challenge on Twitter's Home Timeline](https://mlsanity.com/paper/2004.13715/) (2020)
- [Social Media and Information Overload: Survey Results](https://mlsanity.com/paper/1306.0813/) (2013)
- [Optimizing Revenue while showing Relevant Assortments at Scale](https://mlsanity.com/paper/2003.04736/) (2020)

## Abstract (arXiv)

We address the problem of maximizing user engagement with content (in the form of like, reply, retweet, and retweet with comments)on the Twitter platform. We formulate the engagement forecasting task as a multi-label classification problem that captures choice behavior on an unsupervised clustering of tweet-topics. We propose a neural network architecture that incorporates user engagement history and predicts choice conditional on this context. We study the impact of recommend-ing tweets on engagement outcomes by solving an appropriately defined sweet optimization problem based on the proposed model using a large dataset obtained from Twitter.