arXiv：cs.AI（全量分类）· Yixi Zhou, Sikun Wang, Lei Fan, Fan Zhang·· 5 小时前AI 评分35
EviGraph：基于时序公共服务知识图谱的带证明选择性推荐
EviGraph: Proof-Carrying Selective Recommendation over Temporal Public-Service Knowledge Graphs
AI 导读
EviGraph 通过区分关键决策需求与可悬置信息，用时序知识图谱中的语言智能体关联证据并由确定性检查器验证推荐是否成立，在双语香港公共服务基准上减少了不必要的弃答。研究发现，额外验证可能撤回本已成立的推荐却不提升决策质量，可靠证据导航的关键在于明确决策必须确立什么，而非单纯增加验证。
正文
Abstract:Public-service recommendations require evidence that matches the requested service, scope, and date. Yet treating every missing detail as decisive can withhold useful recommendations. We introduce EviGraph, which distinguishes critical decision requirements from information that can remain unresolved. A language agent links these requirements to evidence in a temporal knowledge graph, while a deterministic checker establishes whether a recommendation is supported. Evaluation on a bilingual Hong Kong public-service benchmark with executable policy references shows that this distinction reduces unnecessary abstention. Additional verification, however, can withdraw supported recommendations without improving decision quality. These findings suggest that reliable evidence-based navigation depends on specifying what must be established for a decision, rather than simply adding more verification.
| Comments: | 19 pages, including figures and tables |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.00212 [cs.AI] |
| (or arXiv:2610.00212v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.00212 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Fan Zhang [view email]
[v1]
Mon, 21 Sep 2026 13:26:48 UTC (392 KB)
来源：arXiv：cs.AI（全量分类） · arxiv.org