2019年8月28-30日，由济南大学主办、山东大学等单位协办的第二届大数据科技国际会议（ICBDT 2019）在济南召开。

瑞典Dalarna University的William Song教授、中国人民大学高瓴人工智能学院经理文继荣教授、山东大学软件学院经理崔立真教授、我司张文宇教授等分别做了大会主旨报告。张文宇教授的报告题目是“Key technologies of stock forecasting models based on big data and crowd intelligence”。

此次会议涵盖了大数据挖掘与算法、数据架构、数据管理、深度学习、大数据可视化、大数据挖掘应用等多个主题，旨在交流大数据领域的最新成果和实践经验。

下附张文宇教授的报告摘要：

    Along with the economic development, an increasing number of factors and high-order data have considerably influenced the fluctuations in stock market. Many existing studies have considered the spatio-temporal correlation of stock index but ignored the comprehensive influence of crowd intelligence on stock index with changes in time and space. Therefore, to improve the forecasting accuracy of the multi-factor and high-order time series stock forecasting models, some key technologies based on big data and crowd intelligence are explored, by combining knowledge graph, crowd sensing, semantic search, personalized recommendation, game theory, fuzzy time series, credit scoring, deep learning, etc. The experimental results demonstrate that the proposed model shows outstanding forecasting accuracy compared with the benchmark methods on the Shanghai Stock Exchange Composite Index and Taiwan Stock Exchange Capitalization Weighted Stock Index.