I Know the Feeling: Learning to Converse with Empathy

Hannah Rashkin, Eric Michael Smith, Margaret Li, Y-Lan Boureau

Abstract: Beyond understanding what is being discussed, human communication requires an awareness of what someone is feeling. One challenge for dialogue agents is recognizing feelings in the conversation partner and replying accordingly, a key communicative skill that is trivial for humans. Research in this area is made difficult by the paucity of suitable publicly available datasets both for emotion and dialogues. This work proposes a new task for empathetic dialogue generation and EmpatheticDialogues, a dataset of 25k conversations grounded in emotional situations to facilitate training and evaluating dialogue systems. Our experiments indicate that dialogue models that use our dataset are perceived to be more empathetic by human evaluators, while improving on other metrics as well (e.g. perceived relevance of responses, BLEU scores), compared to models merely trained on large-scale Internet conversation data. We also present empirical comparisons of several ways to improve the performance of a given model by leveraging existing models or datasets without requiring lengthy re-training of the full model.

Keywords: dialogue generation, nlp applications, grounded text generation, contextual representation learning

TL;DR: We improve existing dialogue systems for responding to people sharing personal stories, incorporating emotion prediction representations and also release a new benchmark and dataset of empathetic dialogues.

Data: [DailyDialog](https://paperswithcode.com/dataset/dailydialog)

Community Implementations: [ 8 code implementations](https://www.catalyzex.com/paper/i-know-the-feeling-learning-to-converse-with/code)