Qi Li (Quinn)
Post-doc
qili5 at illinois dot edu
2128 Siebel Center
Department of Computer Science, UIUC
I am broadly interested in data management with a focus on data mining and machine learning, especially in data aggregation and data quality analysis for truth discovery and crowdsourcing. I have contributed several key methods and is recognized as an expert in these fields.
Truth Discovery
Truth discovery is an emerging research topic in database and data mining communities. The goal of truth discovery is to find the trustworthy information and resolve conflicts among multiple sources without supervision. This goal is usually achieved by jointly estimating source reliability and the trustworthy information. I have been working on several truth discovery projects driven by different applications, such as healthcare and biomedical systems, web mining, wisdom of the crowd, environmental monitoring, smart city, and knowledge base construction. The projects result in over 20 publications and 4 tutorials on highly competitive conferences and journals in data mining and related fields.
Human-Generated Data
Human provides knowledge in various forms. One important form is in text, such as scientific literature, news articles, and reports. I am putting many efforts in extracting structured knowledge from the unstructured text for better knowledge organization, efficient knowledge search, and inspiring new knowledge discovery. My current focus is on the biomedical domain, where the text is abundant but the structured knowledge is still sparse and in huge demand. Another important format of human-generated data is via crowdsourcing, where human workers provide their knowledge on small tasks. I have been working on projects that mainly focus on the following two tasks in crowdsourcing: label aggregation and budget allocation. The mechanisms are designed as general tools to make crowdsourcing easier and more effective. My contribution in crowdsourcing also includes bridging the fields of truth discovery and crowdsourcing.