# Deep Learning for User Comment Moderation

## TL;DR

This paper demonstrates that an RNN with a classification-specific attention mechanism outperforms previous models in user comment moderation tasks using large multilingual datasets.

## Contribution

It introduces a deep RNN model with a novel attention mechanism for improved comment moderation performance.

## Key findings

- RNN outperforms previous state-of-the-art models
- Attention mechanism enhances moderation accuracy
- Model effective on both Greek and English datasets

## Abstract

Experimenting with a new dataset of 1.6M user comments from a Greek news portal and existing datasets of English Wikipedia comments, we show that an RNN outperforms the previous state of the art in moderation. A deep, classification-specific attention mechanism improves further the overall performance of the RNN. We also compare against a CNN and a word-list baseline, considering both fully automatic and semi-automatic moderation.