TRACE: State-Aware Query Processing over Temporal Evidence Graphs for Conversational Data
Quick Answer
TRACE is a novel query processing framework for conversational data that utilizes temporal evidence graphs to enhance state-aware reasoning.
Quick Take
It improves long-conversation query-answering by effectively managing evolving user states, achieving better temporal and multi-hop reasoning in benchmarks. The framework maintains validity annotations to differentiate between current and obsolete facts.
Key Points
- TRACE models conversations as hierarchical graphs with temporal and causal relations.
- The framework separates lexical recall from evidence reconstruction for efficient querying.
- Experiments show improved performance in temporal and multi-hop reasoning tasks.
- Validity annotations allow access to historical facts while prioritizing current information.
- TRACE enhances long-conversation QA benchmarks with a hybrid context for answers.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Conversational data is increasingly used as a persistent source of user state for long-running assistants and AI agents. However, querying this data remains challenging because conversations naturally evolve: plans are revised, preferences change, and later messages frequently supersede or contradict earlier information. Existing long-memory pipelines largely treat memories as independent text or vector objects. This approach often retrieves semantically similar but stale evidence, offering limited support for state-aware reasoning. To address this problem, we present TRACE, a query processing framework over temporal evidence graphs for evolving conversational data. TRACE models conversations as a hierarchical graph spanning events, sessions, and topics, enriched with typed temporal, causal, update, and contradiction relations. Crucially, the framework maintains validity annotations so obsolete facts remain accessible for historical queries but are discounted for current-state answers. At query time, TRACE combines vector-based note retrieval with graph-guided evidence search, generating validity-aware support paths and a hybrid context for answer generation. This design separates lexical recall from evidence reconstruction, enabling bounded query-time reasoning over long conversational histories. Experiments on long-conversation query-answering (QA) benchmarks show that TRACE improves temporal and multi-hop reasoning, with ablations highlighting the importance of hierarchy, update-aware seeding, and path-grounded evidence.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2607.00339 [cs.CL] |
| (or arXiv:2607.00339v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2607.00339 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Maolin Wang [view email]
[v1]
Wed, 1 Jul 2026 02:28:47 UTC (1,750 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CL
See more →TriAgent: Divergence-Aware Committees for Cost-Efficient Financial Sentiment Analysis
TriAgent introduces a cost-efficient multi-agent system for financial sentiment analysis, combining VADER, FinBERT, and Qwen2.5. It achieves an F1 score of ~0.87 with significant savings of $9.3M/year at a 10M-user scale compared to GPT-4o-mini, while also detecting hallucinations with an AUC of 0.90.