jXBW: Fast Substructure Search for Large-Scale JSONL Datasets with LLM Applications

TL;DR

jXBW is a novel compressed index that enables fast, scalable substructure search in large JSONL datasets, significantly improving efficiency for applications like LLM prompts, chemical data, and geospatial analytics.

Contribution

The paper introduces jXBW, a new compressed index with innovative tree representation and search algorithms, enabling efficient substructure search in large JSONL datasets.

Findings

Achieves up to 4,700× speedup over tree-based methods.

Over 6 million times faster than XML-based approaches.

Enables practical large-scale JSONL substructure search.

Abstract

JSON Lines (JSONL) is widely used for managing large collections of semi-structured data, ranging from large language model (LLM) prompts to chemical compound records and geospatial datasets. A key operation is substructure search, which identifies all JSON objects containing a query pattern. This task underpins applications such as drug discovery (querying compounds for functional groups), prompt engineering (extracting prompts with schema fragments), and geospatial analytics (finding entities with nested attributes). However, existing methods are inefficient: traversal requires exhaustive tree matching, succinct JSON representations save space but do not accelerate search, and XML-based approaches incur conversion overhead and semantic mismatches. We present jXBW, a compressed index for efficient substructure search over JSONL. jXBW introduces three innovations: (i) a merged tree representation that consolidates repeated structures, (ii) a succinct tree index based on the eXtended Burrows--Wheeler Transform (XBW), and (iii) a three-phase algorithm for substructure search. These enable query-dependent complexity, where cost depends on query characteristics rather than dataset size, while retaining succinct space. This resolves a key bottleneck in retrieval-augmented generation (RAG) systems requiring structure-aware retrieval. Experiments on seven real datasets, including PubChem (1M compounds) and OSM geospatial data (6.6M objects), achieve up to 4,700$\times$ speedup over tree-based methods and over $6\times 10^6$ speedup relative to XML-based approaches. jXBW makes JSONL substructure search practical for the first time, opening opportunities for large-scale LLM-based analytics.