Table Mining Initiative

Intelligent Table Reasoning for AGI Research

The USTC AGI Table Mining Group designs benchmarks, toolchains, and autonomous agents that push the boundaries of table understanding, time-series reasoning, and multimodal analytics.

Research Focus Table OCR, Temporal Table Classification, Table Reasoning, Table Agents

Research Pillars

From Recognition to Autonomous Table Agents

We connect perception, language, and reasoning to make tables first class citizens for AGI systems.

Table Recognition

NGTR benchmark plus Neighbor-Guided Toolchain Reasoner improve table structure parsing for low-quality inputs.

NGTR · IJCAI 2025 → GitHub →

Temporal Table Intelligence

TableTime reformulates multivariate time-series classification as table understanding with training-free LLM reasoning.

TableTime · CIKM 2025 → GitHub →

Slow Table Reasoning

STaR develops slow-thinking LLMs with uncertainty-aware inference for stable multi-hop table reasoning.

STaR · Preprint →

Comprehensive Survey

A sweeping review of table mining with LLMs surfaces challenges, advances, and future opportunities.

Table Mining Survey →

Flagship Works

Benchmarks, Frameworks, and Agents

2026

TableMind · WSDM

Autonomous programmatic agent with tool-augmented reasoning, RAPO training, and sandboxed execution for reliable analytics.

Paper →

2025

NGTR · IJCAI

Hierarchical benchmark plus neighbor-guided toolchain to fix low-quality table images and boost VLLM recognition.

Paper →

2025

TableTime · CIKM

Converts multivariate time series to textual tables, enabling zero-shot classification with LLM reasoning pipelines.

Paper →

2024+

PoTable & STaR

Stage-oriented plan-execute agents and slow-thinking LLMs create transparent, uncertainty-aware reasoning trajectories.

Resources

Toolchains, Data, and Guidance

NGTR Benchmark Suite

Structured evaluation pipeline, visual toolchain, and reflection module for robust table recognition.

TableTime Reasoning Stack

Prompt templates, neighborhood assistance, and multi-path inference ready for zero-shot classification.

Agentic Tool Libraries

Plan-then-execute operation sets with code generation, sandbox execution, and feedback monitoring.

ChemTable Dataset

Expert-curated chemical tables with cell polygons, logical layouts, and domain labels.

Why Table Mining Now?

Tables are the connective tissue of scientific reports, enterprise systems, and financial decision making. We elevate tables to become native reasoning artifacts for AGI, empowering models with precise structure, semantic context, and executable toolchains.

Human-AI Collaboration

Multi-stage workflows provide interpretable reasoning plans and verifiable outputs for analysts.

Simulation to Deployment

Benchmarks mirror messy real-world tables so agents stay robust beyond curated datasets.

Open Science

We share toolchains, prompts, and datasets to accelerate community progress in structured reasoning.