报告人简介：

赵洪科，现为天津大学管理与经济学部教师，CCF会员、CAAI会员。2018年11月毕业于中国科学技术大学，获计算机科学与技术专业博士学位。曾分别于2015年9月－2016年8月在美国北卡罗来纳大学夏洛特分校计算与信息学院，2016年8月－2017年1月在美国亚利桑那大学埃勒商学院信息管理系统系等知名高校访问留学。研究方向包括数据挖掘、机器学习，特别是人工智能技术在金融商务大数据和Fintech等领域的应用。在计算机、信息管理等领域的高水平学术期刊（如TKDE, TIST, TSMC, TBD, IMM, Information Sciences, Scientometrics等）和顶级国际学术会议（如SIGKDD，AAAI，IJCAI，ICDM等）上发表论文30余篇。获中国人工智能学会（CAAI）优秀博士学位论文提名奖、中国科学技术大学优秀博士学位论文奖、中国科学院院长奖、中国科学院朱李月华优秀博士生奖、中国机器学习会议（CCML）最佳学生论文奖等奖项。

报告简介：

In this talk, I will present our recent work on exploring the knowledge creation theory (SECI) in management with machine learning. Conventionally, knowledge creation theories mainly elaborate the offline context, not considering strengthened membership fluidity or rapid trust in online social Q&A sites, which makes it urgent to construct a new theory to explain the phenomenon. However, fluid and heterogeneous knowledge creation in social Q&A sites are pressing to describe, so we designed a special deep learning model to quantify process to construct a process model based on concepts of socialization, externalization, combination and internalization in SECI theory. Surprisingly, results show strength of knowledge creation could change with time and there are correlations of specific four processes. This pattern expands traditional knowledge creation theory into a dynamic process including longitudinal and transverse change, contributing largely to knowledge creation theory and provides a quantitative methodology to explore theories.