AI for accountable business
Natural language processing, computer vision, and multimodal AI that preserve provenance and support decisions people can defend.
AI · NLP · Computer visionResearcher · Educator · Applied AI Builder
AI that earns trustin real decisions.
I study how artificial intelligence, narrative data, and human judgment can work together to improve corporate disclosure, sustainable business, and learning.
让AI不止生成答案，而是形成可信、可执行、可追责的决策。
Assistant ProfessorHead, Business Analytics Research ClusterXi’an Jiaotong-Liverpool University
Current focusMultimodal AI × ESG evidence
RMB 300k+external industrial funding
9research & collaborative projects
FHEA · CPAacademic and professional recognition
Research agenda
From prediction
to accountable action.
The technical baseline is becoming abundant. My work concentrates on what remains scarce: defensible measures, reliable evidence, domain judgment, and adoption inside real institutions.
ESG & sustainable finance
Turning disclosure narratives and visual evidence into measurable signals of corporate behavior, supply-chain practice, and value.
ESG · Disclosure · Sustainable supply chainsHuman–AI learning
Designing project-based learning where AI scaffolds inquiry while teachers retain judgment, evidence standards, and responsibility.
PBL · AI governance · Learning designSelected publications
Research built to be reused.
01
ESG · NLP · Sustainable supply chains
2025↗02Decoding ESG Report Narratives: Unveiling Sustainable Supply Chain Insights and Impacts Through Textual Analysis
Corporate Social Responsibility and Environmental Management, 32(2), 2559–2581
Tax avoidance · Earnings quality
2026↗03Non-conforming Tax Avoidance and Real Earnings Management
Asian Review of Accounting, 1–29
Tax strategy · Earnings persistence
2025↗Conforming Tax Avoidance and Earnings Persistence
Pacific Accounting Review, 37(1), 89–112
Translation & impact
Research leaves the page.
I work at the boundary between academic evidence and operating reality—where a model must survive data limitations, organizational incentives, and the consequences of a decision.
Teaching & learning
AI should deepen inquiry,
not bypass it.
My teaching combines project-based learning, authentic industry problems, and human–AI collaboration. Students learn to make claims, test evidence, and defend decisions—not merely produce polished outputs.
Data Mining ApplicationsInformation Systems ManagementSyntegrative Education in Deep Practice
2026
Syntegrative Education Excellence Award
2025
XEC Champions Best Practices Sharing Award
2024
Teaching Excellence Award
2024
Excellent Supervisor, Citi Cup Fintech Competition
2021 & 2020
Monash Purple Letter for Outstanding Teaching Performance
Research pipeline
What I am building next.
Signaling or Obfuscating? Green Bond Issuance and Visual–Textual Consistency in ESG Disclosures
A multimodal inquiry into whether what companies show aligns with what they say in ESG reporting.
Provenance-preserving Multiview AI for ESG Disclosure
Developing auditable methods that integrate text, visuals, external evidence, and stakeholder perspectives without erasing source provenance.
Open to serious collaboration
Bring a consequential
question.
Research collaboration, doctoral supervision, responsible AI, ESG analytics, and evidence-led industry projects.
Start a conversation ↗