# What You Say and How You Say it: Joint Modeling of Topics and Discourse in Microblog Conversations

## TL;DR

This paper introduces an unsupervised neural framework that jointly models topics and discourse in microblog conversations, improving understanding and classification of message content and behavior.

## Contribution

It proposes a novel neural model for unsupervised joint learning of topics and discourse, enhancing interpretability and classification performance in microblog conversations.

## Key findings

- The model produces coherent topics and meaningful discourse representations.
- Joint training improves microblog message classification accuracy.
- The approach outperforms existing methods in capturing conversation dynamics.

## Abstract

This paper presents an unsupervised framework for jointly modeling topic content and discourse behavior in microblog conversations. Concretely, we propose a neural model to discover word clusters indicating what a conversation concerns (i.e., topics) and those reflecting how participants voice their opinions (i.e., discourse). Extensive experiments show that our model can yield both coherent topics and meaningful discourse behavior. Further study shows that our topic and discourse representations can benefit the classification of microblog messages, especially when they are jointly trained with the classifier.