WikiKV: Schema-Evolving Path-Indexed Storage for Hierarchical Knowledge Navigation

2026-06-12

AI Summary

Existing storage systems struggle to efficiently support knowledge base workloads that are hierarchical, subject to high query frequency, and continuously evolving. This work proposes a specialized storage system tailored for tree-structured wiki knowledge bases used in large language model (LLM) applications. The design introduces several key innovations: path-indexed key-value storage, intent-anchored schema induction, continuous evolution operators, a lock-free consistency protocol, and budget-guided navigation. These mechanisms collectively reduce LLM invocation frequency, ensure progressively refined answers, and enable low-latency, balanced query performance. Evaluated on the AuthTrace dataset, the system achieves an end-to-end answer accuracy of 63.2%, substantially outperforming various retrieval-augmented generation (RAG) baselines, particularly in multi-document scenarios with heterogeneous fan-in structures.

Abstract

LLM-curated hierarchical knowledge bases, namely a tree-structured wiki whose nodes summarize an underlying corpus, have become a dominant substrate for retrieval-augmented applications, yet their storage layer is still treated as an implementation detail. This workload is hierarchical, query-intensive, and continuously evolving, and no existing storage model natively captures all three properties at once. We present WikiKV, a path-indexed key-value storage model purpose-built for this workload, comprising three components: (i) a data-driven schema that bootstraps the hierarchy via Intent-Anchored Schema Induction and refines it through Continuous Evolution Operators; (ii) a consistency protocol for the path-indexed storage model that precludes partial-read observations under concurrent offline rewrites without read-path locking; and (iii) a budgeted navigation operator whose search-accelerated routing reduces the expected number of LLM-assisted descent steps from d to O(1) while preserving anytime semantics with progressively refined answers. We evaluate WikiKV through real-world deployment for the WeChat Official Account AI Assistant and benchmark it against diverse baselines on the AuthTrace dataset, where it achieves balanced low per-operator latency across four query operators against relational, graph, and FS backends, and reaches 63.2% end-to-end answer correctness, exceeding multiple RAG baselines, with the gap widening on low- and high-fan-in multi-document questions. Ablation study further confirms the effectiveness of WikiKV's components.

Problem

• hierarchical knowledge base
• schema evolution
• query-intensive workload
• path-indexed storage
• retrieval-augmented generation

Innovation

• path-indexed storage
• schema evolution
• hierarchical knowledge navigation
• consistency protocol
• budgeted navigation