Knowledge Base assisted Text Categorization
Evan Wei Xiang, Qiang Yang · Rare & Special e-Zone (The Hong Kong University of Science and Technology) · 2009
With the development of Web 2.0, people are more willing to retrieve and share through online knowledge base. Such information source contains rich human annotated information. In this paper, we will summarize our current research works on how to tackle the difficulties of traditional data mining tasks with the knowledge retrieved from online knowledge bases. We use the sparse text categorization and domain adaptation as case study problems and Wikipedia as knowledge base to demonstrate how we achieve the knowledge transfer via feature projection and instance migration.