Learning Latent Local Conversation Modes for Predicting Community Endorsement in Online Discussions
read the original abstract
Many social media platforms offer a mechanism for readers to react to comments, both positively and negatively, which in aggregate can be thought of as community endorsement. This paper addresses the problem of predicting community endorsement in online discussions, leveraging both the participant response structure and the text of the comment. The different types of features are integrated in a neural network that uses a novel architecture to learn latent modes of discussion structure that perform as well as deep neural networks but are more interpretable. In addition, the latent modes can be used to weight text features thereby improving prediction accuracy.
This paper has not been read by Pith yet.
Forward citations
Cited by 1 Pith paper
-
KARMA: Karma-Aligned Reward Model Adaptation
KARMA adapts reward models from Reddit karma data to align LLMs with conversational pragmatics, finding that context-only rewards outperform karma-predictive ones downstream while reducing factuality across conditions.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.