Learning Structured Knowledge from Real-World Data without Excessive Annotations

Enriching Open-Domain Dialogue Models with Predictive Social Commonsense

Leveraging Diverse Data Generation for Domain-Adaptable Dialogue State Tracking

Automating Biomedical Abstract Screening using Network Embedding

Contextual Embedding Representations for Retrieval-Based and Generation-Based Dialogue Systems

Attention-enhanced Deep Learning Models for Data Cleaning and Integration

Computational Structures as Neural Symbolic Representation

Overcoming Spatiotemporal Trade-offs in Calcium Imaging Using Deep Learning-based Dynamics Modeling

Knowledge-Aware User Intent Inference for Web Search and Conversational Agents