The goal of my research is to unleash hidden knowledge buried in unstructured text data and enable them to be more accessible, interpretable, and reusable. Towards this goal, I propose a novel and generic data model, named multi-faceted taxonomy, which organizes concepts of different facets into hierarchical structures. My research consolidates the power of multi-faceted taxonomy in three areas of investigation:
- Construction: To identify important concepts and their taxonomic relations from text corpora, I propose a series of set expansion and topic/word hierarchy construction methods.
- Enrichment: To keep existing taxonomies up-to-date in real-world applications, I study multiple tasks including synonym (set) discovery, taxonomy expansion, and taxonomy completion.
- Application: To distill knowledge from multi-faceted taxonomies for downstream applications, I develop methods for weakly-supervised text classification and unsupervised literature search.
UIUC CS512: Data Mining Principle
Teaching Assistant • Spring 2020
UIUC CS412: Introduction to Data Mining
Guest Lecturer • Spring 2019
SJTU CS119: Principles of Computer Algorithms
Teaching Assistant • Summer 2015
SJTU CS114: Programming and Data Structures
Teaching Assistant • Spring 2015
SJTU MA118: Mathematical Analysis I
Teaching Assistant • Fall 2013