AI-generated evidence summaries yield crowd truthfulness judgments with accuracy comparable to full webpages while cutting time by roughly 13-15 percent.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Efficiency and Effectiveness of LLM-Based Summarization of Evidence in Crowdsourced Fact-Checking
AI-generated evidence summaries yield crowd truthfulness judgments with accuracy comparable to full webpages while cutting time by roughly 13-15 percent.