.png){kind=icon}
.png){kind=icon}
Return
Adaptive complex knowledge base question answering with program decomposition chain
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2025.131736.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose PDC-ACQA, an effective adaptive complex question answering framework for decomposable and non-decomposable queries. • We design a structural program decomposition chain that represents the overall progressive decomposition process of queries. • We introduce an expert feedback-based self-correction strategy to evaluate and refine each step’s decomposition result. • Experimental results demonstrate that PDC-ACQA effectively improves the overall performance of PI-based question answering.
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W