您好,欢迎访问东北财经大学管理科学与工程学院!(请使用IE8及以上浏览器访问,360浏览器请开启极速模式) English| 中文| 加入收藏

电子商务系列研讨会(18):Data Mining Applications in Business Research

    为进一步深入学术交流,增进学术氛围,应电子商务系邀请,美国弗吉尼亚理工大学博士生杜前舟将于2017年7月10日来我院进行学术交流并作学术报告。报告将结合数据挖掘技术在商业领域的应用现状及其最新研究做题为“Data Mining Applications in Business Research”的报告。欢迎感兴趣的教师及研究生参加。
 
    报 告 人:杜前舟
    会议时间:2018年7月10日 8:00-9:30
    会议地点:劝学楼425
 
【报告人简介】
    杜前舟,美国弗吉尼亚理工大学潘普林商学院商业信息技术专业在读博士生。主要研究领域包括数据挖掘、文本分析、社交媒体分析、大众智慧、商业智能技术、决策分析等。相关研究成果发表在《Journal of Management Information Systems》、《Tourism Management》、《International Journal of Hospital Management》、《Information Technology & Tourism》等国际重要期刊上。
 
 
Abstract
Part 1: Data Mining Applications in Business Research
 Abstract
 Data mining techniques have been applied in multiple business research areas, such information systems, tourism management, accounting, finance, and marketing. Due to the increasing availability of big data, effectiveness and scalability of information technology, data mining methods furtherly become popular. By now, data mining models can be used to automatically generate or estimate variables that researchers are interested in. For example, based on these models, we can extract some useful variables, such as sentiment and semantics of text, travel purpose, image content, and individual opinion. In this talk, I will introduce some useful techniques, which were applied in my previous work, to extract some interesting variables.
 Keywords: data mining techniques, business research
 
Part 2: Opinion Aggregation from Social Crowds for Stock Prediction
Abstract
The Wisdom of the Crowds theory explains the phenomenon in which the aggregated opinion of a diverse group of individuals is closer to the truth than most individuals in the group. Guided by the theory, we design an opinion aggregation method, namely Social Crowd IQ (SCIQ), for extracting wisdom from social crowds to predict stock trends. We include in our design three features in order to account for the issues possibly impacting the wisdom of social crowds. We use controlled experiments to show that SCIQ outperforms the baseline opinion aggregation methods in terms of the crowd performance for stock prediction. In addition, we conducted functional testing to show that each of the three design elements is important in achieving the best crowd prediction performance. A trading strategy powered by SCIQ achieved 6% more return than that driven by the best baseline opinion aggregation method in a simulated investment scenario.  
Keywords: Wisdom of the crowds, online investment community, opinion aggregation, FinTech, stock prediction
 
 
撰稿人: 宋晓龙                                审核:田甜