How did the discussion go: Discourse act classification in social media conversations

TL;DR

This paper introduces a hierarchical LSTM model with attention for classifying discourse acts in social media conversations, demonstrating improved domain-independent performance across Reddit and Facebook datasets.

Contribution

The study presents a novel attention-based hierarchical LSTM approach for discourse act classification that outperforms previous models without relying on platform-specific features.

Findings

Achieved F1-scores of 71% on Reddit and 66% on Facebook datasets.

Model outperforms previous approaches in domain independence.

Attention mechanism reveals relevance of text segments in discourse roles.

Abstract

We propose a novel attention based hierarchical LSTM model to classify discourse act sequences in social media conversations, aimed at mining data from online discussion using textual meanings beyond sentence level. The very uniqueness of the task is the complete categorization of possible pragmatic roles in informal textual discussions, contrary to extraction of question-answers, stance detection or sarcasm identification which are very much role specific tasks. Early attempt was made on a Reddit discussion dataset. We train our model on the same data, and present test results on two different datasets, one from Reddit and one from Facebook. Our proposed model outperformed the previous one in terms of domain independence; without using platform-dependent structural features, our hierarchical LSTM with word relevance attention mechanism achieved F1-scores of 71\% and 66\% respectively to predict discourse roles of comments in Reddit and Facebook discussions. Efficiency of recurrent and convolutional architectures in order to learn discursive representation on the same task has been presented and analyzed, with different word and comment embedding schemes. Our attention mechanism enables us to inquire into relevance ordering of text segments according to their roles in discourse. We present a human annotator experiment to unveil important observations about modeling and data annotation. Equipped with our text-based discourse identification model, we inquire into how heterogeneous non-textual features like location, time, leaning of information etc. play their roles in charaterizing online discussions on Facebook.