Generative agents built from a Filipino voter survey reproduce and simulate sentiment ratings with 81% to 92% weighted agreement, but the target attitudes may already be embedded in the agent profiles.
Mining the peanut gallery: Opinion extraction and semantic classification of product reviews
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Sentiment Simulation using Generative AI Agents
Generative agents built from a Filipino voter survey reproduce and simulate sentiment ratings with 81% to 92% weighted agreement, but the target attitudes may already be embedded in the agent profiles.