摘要
随着大数据时代的到来,信息系统收集了海量情境信息,如舆情信息、环境信息、经济信息等。这些情景大数据提供丰富的细节信息,更细致地刻画行为背景以辅助用户行为建模。阐述了两种使用表达学习策略建模一般化情境信息的框架,并针对情境大数据中最常见的时序情境建模问题,使用循环神经网络建模时序情境中的序列依赖关系。
In the big data era, information system has to handle a mass of data of contextual information, such as public opinion, environment information and economic status. Embedded with abundant details of user behavior, contextual information plays a significant role in effectively shaping user character and elaborately modeling user behavior. Two frameworks to model general context information through representation learning and a recurrent model for sequential context scenarios were involved.
作者
吴书
刘强
王亮
WU Shu LIU Qiang WANG Liang(Center for Research on Intelligent Perception and Computing, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China)
出处
《大数据》
2016年第6期110-117,共8页
Big Data Research
基金
国家自然科学基金资助项目(No.61403390
No.U1435221)~~
关键词
情境大数据
情境建模
用户建模
行为预测
contextual big data
context modeling
user modeling
behavior prediction
