Skip to content
Open access

ACMO: An Adaptive Cost Modeling and Optimization Framework for Schema-Heterogeneous NoSQL Document Databases

2026 · IEEE Access · Vol 14, pp. 162489-162508 · 0 citations · 17 references

TL;DR

ACMO, the Adaptive Cost Modeling and Optimization framework, is introduced, which estimates query costs on MongoDB-style document collections without requiring query execution traces or a running database engine, relying instead on document structure and field-level statistics.

Abstract

Query cost estimation in schema-heterogeneous NoSQL document databases remains an open problem, since collections with zero structural entropy render classical entropy-based cost signals uninformative, while progressive type drift invalidates the stationarity assumptions of fixed-schema learned models. This paper introduces ACMO, the Adaptive Cost Modeling and Optimization framework, which estimates query costs on MongoDB-style document collections without requiring query execution traces or a running database engine, relying instead on document structure and field-level statistics. We evaluate ACMO against seven baselines—four document-domain models and three learned ensemble approaches—on five real-world MongoDB Atlas collections (5,000–50,000 documents) spanning nested, flat, schema-drift, and mixed structural regimes. ACMO achieves the lowest mean MAE (3,282 ms), ahead of the strongest gradient-boosted competitor, CatBoost (3,491 ms), while structural and analytical baselines trail by an order of magnitude. Paired significance testing confirms that ACMO’s advantage over the weaker baselines is statistically decisive in the pooled comparison, with no significant difference from the gradient-boosted models on any individual collection—indicating comparable accuracy at a fraction of the model complexity. We further validate ACMO’s cost ordering against live, indexed MongoDB execution, obtaining significant rank correlation on four of five collections and identifying, with root-cause analysis, the specific limitation underlying the fifth. ACMO trains in under 2.7 seconds without database connectivity, suiting cold-start and post-migration scenarios.

Read PDF

Similar papers

Open access Oct 2026

Structured AI: A Three-Tier Adaptive Framework for Enterprise Data Analytics

Traditional Retrieval-Augmented Generation (RAG) systems and SQL-based query engines face significant challenges when processing real-world Excel files containing multiple tables, merged cells, multi-row headers, formulas, pivot-style layouts, and inconsistent schemas. Structured AI is a hybrid three-tier framework tha...

B. Mohanty, Gopichand Agnihotram, Raja Sekhar Reddy S · 0 citations
Open access Oct 2026

Performance Evaluation of Relational, JSON, and Hybrid Models in MariaDB for Dynamic Form Builders

Dynamic form builders require database structures that can accommodate variable attributes while maintaining efficient query performance. However, the performance implications of different storage representations within the same database management system remain insufficiently characterized, particularly for relational...

Muhammad Salmansyah Mousavi Akhyar, Triyanna Widiyaningtyas · 0 citations
Preprint Oct 2026

Prune First, Decide Fast: Scalable Semantic Query Processing with JEVDB

JEVDB is presented, a scalable semantic database system that uses fast, typed decision models for semantic filters, joins, classification, and ranking, while selectively escalating uncertain cases to generative LLMs.

Zheng-Le Wang, Han-Xu Yan, Fuheng Zhao et al. · 3 citations · ⚡1
Open access 2025

CODAQ: Context-Oriented Distributed Adaptive Query Processing for Large-Scale Graph and Data Systems

Modern data-driven applications increasingly rely on distributed graph databases and machine learning pipelines to extract knowledge from massive datasets. However, efficient query processing in distributed environments remains challeng-ing due to issues such as graph partitioning inefficiencies, dynamic connectivity u...

Praveen Kumar Kumbum, Ramachandra Reddy Vangala · 4 citations
#natural language process... Preprint Oct 2026

Schema-Free Text-to-SQL: Learning Enterprise Database Structure from Query Logs

Text-to-SQL systems achieve remarkable accuracy on academic benchmarks, yet performance drops dramatically in enterprise deployments. This gap stems from a fundamental assumption: existing approaches require complete schema documentation, but enterprise databases typically lack comprehensive documentation, featuring cr...

Ömer Faruk Akgül, Nikhil Kapahi, P. Muppidi · 0 citations
Review Open access Oct 2026

Next-Generation Query Optimization: A Systematic Review of Cost Models, Equivalence Rules, and Materialized Views

The evidence indicates that adaptive cost models achieve 15–30% efficiency gains over classical models but at higher computational cost, graphbased equivalence transformations reduce query latency by around 18% compared to static rules, and materialized views deliver 15–32% performance gains but introduce 10–20% storag...

M. Okoronkwo · 1 citation

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.