BeamAggR: Beam Aggregation Reasoning over Multi-source Knowledge for Multi-hop Question Answering
This addresses retrieval and integration challenges in multi-hop QA for AI systems, representing an incremental advance in reasoning frameworks.
The paper tackles the problem of factual errors in large language models for knowledge-intensive multi-hop question answering by proposing BeamAggR, a reasoning framework that parses questions into trees and uses probabilistic aggregation to prioritize answers, resulting in a 8.5% improvement over state-of-the-art methods on four datasets.
Large language models (LLMs) have demonstrated strong reasoning capabilities. Nevertheless, they still suffer from factual errors when tackling knowledge-intensive tasks. Retrieval-augmented reasoning represents a promising approach. However, significant challenges still persist, including inaccurate and insufficient retrieval for complex questions, as well as difficulty in integrating multi-source knowledge. To address this, we propose Beam Aggregation Reasoning, BeamAggR, a reasoning framework for knowledge-intensive multi-hop QA. BeamAggR explores and prioritizes promising answers at each hop of question. Concretely, we parse the complex questions into trees, which include atom and composite questions, followed by bottom-up reasoning. For atomic questions, the LLM conducts reasoning on multi-source knowledge to get answer candidates. For composite questions, the LLM combines beam candidates, explores multiple reasoning paths through probabilistic aggregation, and prioritizes the most promising trajectory. Extensive experiments on four open-domain multi-hop reasoning datasets show that our method significantly outperforms SOTA methods by 8.5%. Furthermore, our analysis reveals that BeamAggR elicits better knowledge collaboration and answer aggregation.