Deep Learning for User Comment Moderation
This work addresses content moderation for online platforms, but it is incremental as it builds on existing methods with new data and minor enhancements.
The researchers tackled the problem of user comment moderation by experimenting with a new dataset of 1.6M Greek news comments and existing English Wikipedia datasets, showing that an RNN outperforms previous state-of-the-art methods, with a deep attention mechanism further improving overall performance.
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.