Year: 2025
Type: article
Abstract: This paper aims to explore methods for fact verification of user answers in the community question-and-answer domain. A novel Gated Recurrent Unit (OE-GRU) model based on global evidence information is proposed. The model improves performance by introducing global evidence information and a scaling factor, and designs a general framework for fact verification in community Q&A. Experiments conducted using the AnswerFact dataset compare the performance of existing CNN, RNN, LSTM, GRU, and the improved GRU models. Results show that the OE-GRU model achieves the best accuracy and F1 scores. Additionally, ablation experiments and comparisons with different scale values and pooling methods further validate the effectiveness of the proposed model's components. Overall, this study provides new methods and technical support for fact verification in the community Q&A domain.
Author: Jiankang Han
Institution: North China Electric Power University
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4407999375