NEXUS Lab
We are an interdisciplinary research lab at the intersection of marketing, artificial intelligence, and decision science, focused on building intelligence for the digital and global economy. Our work spans three core areas—marketing intelligence, generative search optimization, and supply chain intelligence—with a shared goal of helping organizations make smarter, faster, and more adaptive decisions. By integrating causal methods, structural modeling, machine learning, multimodal and agentic AI, and network analytics, we turn complex data from consumers, platforms, digital content, and global trade systems into measurable insights and practical decision tools. We seek to advance both scientific understanding and real-world impact, enabling firms to evaluate marketing effectiveness, strengthen brands, optimize search-era visibility, and build resilience in increasingly interconnected markets.
Research Areas
Figure: Automatically screening safety risks from consumer complaints
Figure: Segmenting vending markets based on consumer location-product purchase data
Figure: How the characteristics of topic consumptions and contributions evolve over 40 months on Zhihu
Figure: Building brands in the Digital Age
Figure: Designing Prompt for narrative content marketing using LLMs
Figure: Generating Desired Visual Ads with Stable Diffusions
Figure: Generating Controllable Visual Stimulus with Stable Diffusions
From causal measurement to generative AI: building the future of marketing intelligence
We develop analytical and AI-based approaches to help firms understand consumers, evaluate marketing actions, and design more effective strategies. Our work spans marketing effectiveness, big platform data analytics, branding and global strategy, and generative AI for decision science. By integrating causal methods, econometric modeling, and modern AI tools, we transform complex, unstructured data into actionable insights for business decisions.
Making Visibility in Generative Search Analyzable, Diagnosable, Optimizable, and Measurable
Generative search is reshaping how users discover information, how brands gain exposure, and how content drives business value. GEO Intelligence focuses on how generative engines retrieve, interpret, rank, generate, and cite web content under different query intents, model architectures, and grounding constraints—and, more importantly, what optimization strategies can systematically improve visibility, citation performance, and answer contribution.
Decoding the structure, players, and dynamics of global trade networks
We investigate how global trade networks form, fracture, and adapt under the pressures of geopolitical realignment and technological disruption. We identify key players across both macro and micro dimensions, e.g., the structural roles that countries and firms occupy within trade networks and hidden operational risks from multi-tier dependencies. These network insights feed into a creditworthiness evaluation framework grounded in actual supply chain topology rather than financial statements alone.