Knowledge-driven practice for the capstone project course of data science

Abstract

The rapid development of data applications poses severe challenges as well as significant opportunities for data science talent cultivation. On the demand for data analysis specialty, in this poster the authors report on some preliminary results of a knowledge-driven practice for the capstone project course of data science. The practice consists of the tier of knowledge generalization and specialization and the tier of data science practice. Three kinds of knowledge, including general knowledge from the perspectives of data science, domain knowledge from real-world applications, and engineering expertise knowledge for conducting data analysis process, are refined to implement data science practice to achieve solutions effectively and efficiently for real-world applications. Preliminary results indicate that knowledge-driven practice promotes students to improve the ability of data analysis efficiently, and students attending the course win world-class data analysis competitions, such as KDD (Knowledge Discovery and Data Mining) Cup and Kaggle.