Tingting Wang

About

I obtained my Ph.D. in the School of Computing Technologies from RMIT University in 2025, supervised by Prof. Zhifeng Bao, Prof. J. Shane Culpepper, and Dr. Shixun Huang. My research lies at the intersection of data management and applied AI.

My work focuses on target-driven data curation and analytical pipelines that discover, integrate, select, and orchestrate heterogeneous and multimodal data, enabling reliable and scalable AI systems for cross-disciplinary knowledge discovery and decision-making.

Broadly, I am interested in unifying database principles with modern AI techniques: using AI to make data management systems more adaptive, and using data management foundations to improve the reliability, efficiency, and transparency of AI systems.

Education

Honors & Grants

  • 2026 UQ Centre for Enterprise AI Seed Funding
  • 2022 - 2024 CSIRO Data61 PhD Scholarship
  • 2017 Outstanding Graduate Award of SCU
  • 2016 National Scholarship of China

Recent News

Our work has been accepted by The VLDB Journal.

My project received UQ Centre for Enterprise AI Seed Funding, with thanks for their support.

We released a new survey, A Survey of LLM-Powered Question Answering Through the Lens of Query Processing, together with a curated GitHub repository. The survey introduces a pipeline-oriented taxonomy for LLM-powered QA, organizing recent work into five query-processing stages and highlighting open research gaps for next-generation QA systems. GitHub

Publications

Conference

Journal

Projects

Target-Driven Data Curation

This topic focuses on discovering, assembling, and acquiring high-quality data from heterogeneous data lakes for target-oriented AI applications.

Dataset Discovery

Unified Data Discovery across Query Modalities and User Intents

Unified Data Discovery framework

Modern data lakes support diverse applications such as question answering and fact verification, yet existing data discovery methods are typically designed for a single query modality or specific user intent. This work studies how to uniformly retrieve relevant tables given either natural language statements or tables as queries, without relying on intent-specific modeling. To this end, a unified cross-modal graph learning framework is introduced to learn shared representations for queries and tables by leveraging heterogeneous contextual signals in data lakes under limited supervision, enabling flexible and scalable relevance assessment.

Paper

Dataset Assemblage

Distinctiveness Maximization in Datasets Assemblage

Distinctiveness Maximization framework

Modern data lakes offer abundant datasets, yet existing discovery methods typically evaluate datasets individually, leading to redundant acquisitions and inefficient budget usage. This work studies how to assemble a set of datasets under a budget to maximize the amount of distinct information with respect to a query set. We formulate this as a distinctiveness maximization problem and develop an efficient ML-based method to estimate dataset distinctiveness, enabling scalable greedy selection without expensive exact computation.

Paper Code

Data Point Selection

Optimizing Data Acquisition to Enhance Machine Learning Performance

Data acquisition for machine learning framework

Modern AI depends heavily on training data quality, yet simply collecting more data does not guarantee better performance. This work studies how to select high-quality labeled data from large, heterogeneous data pools to improve a target ML model. Since validating data value typically requires repeated retraining during selection, which is computationally expensive, we develop adaptive, model-agnostic algorithms that combine online learning with dynamic exploration-exploitation scoring to efficiently estimate data utility without full retraining.

Paper Code

Students

Ph.D. Students

Jiyeon Beack, The University of Queensland, Australia Advisor

Jiyun Xu, Wuhan University of Technology, China

Master's Students

Gang Hua, Wuhan University of Technology, China

Xinyu Wang, Nanyang Technological University, Singapore

Honours Student

Pham Hung Cuong Le, The University of Queensland, Australia

Professional Service

Area Chair

KDD 2026

Senior Program Committee Member

PAKDD 2026

Program Committee Member

WWW 2026; CIKM 2023, 2024, 2025

Session Chair

WWW 2025; ADC 2023