Complaint Identification in Social Media with Transformer Networks
This work addresses the need for more accurate complaint detection in social media, which is useful for applications like customer service and sentiment analysis, but it is incremental as it adapts existing transformer models with additional linguistic features.
The paper tackled the problem of automatically identifying complaints in social media by evaluating transformer-based neural models combined with linguistic information, achieving a macro F1 score of up to 87% and outperforming previous state-of-the-art methods.
Complaining is a speech act extensively used by humans to communicate a negative inconsistency between reality and expectations. Previous work on automatically identifying complaints in social media has focused on using feature-based and task-specific neural network models. Adapting state-of-the-art pre-trained neural language models and their combinations with other linguistic information from topics or sentiment for complaint prediction has yet to be explored. In this paper, we evaluate a battery of neural models underpinned by transformer networks which we subsequently combine with linguistic information. Experiments on a publicly available data set of complaints demonstrate that our models outperform previous state-of-the-art methods by a large margin achieving a macro F1 up to 87.