| 摘要: |
| 随着移动互联网的发展与手持智能终端的普及，海量带有用户时空属性的数据被生成.理解这些数据表达的语义信息对推测用户需求，分析用户偏好，进而提供精准时空推荐和预测服务具有重要作用.因此，近些年来，时空数据语义理解正成为时空数据挖掘领域的研究热点.从技术和应用两个层面，对近些年来国内外研究者在该领域的研究成果进行了系统的归类和总结.技术层面上，依据语义理解的不同任务，提出了时空数据语义理解的研究框架；并依次从地理位置语义理解、用户行为语义理解、热点事件语义理解3个主要任务，归纳了时空数据语义理解所包含的相关研究成果和关键技术.应用层面上，分别总结了时空数据语义理解在时空推荐和时空预测中的应用.最后，从数据质量、算法模型和计算模式3个方面，归纳了时空数据语义理解面临的主要挑战以及未来的研究方向. |
| 关键词: 时空数据挖掘 语义理解 语义偏好 用户画像 机器学习 |
| DOI：10.13328/j.cnki.jos.005576 |
| 分类号: |
| 基金项目:国家自然科学基金（61472403，61303243，61702470） |
|
| Semantic Understanding of Spatio-Temporal Data: Technology & Application |
|
YAO Di1, ZHANG Chao2, HUANG Jian-Hui1, CHEN Yue-Xin3, BI Jing-Ping1
|
|
1.Institute of Computing Technology, The Chinese Academy of Sciences, Beijing 100190, China;2.University of Illinois at Urbana-Champaign, US;3.National Key Laboratory of Science and Technology on Blind Signal Processing, Chengdu 610041, China
|
| Abstract: |
| With the development of mobile internet and widespread use of mobile phones, a large amount of data that contains user' time and space attributes has been generated and collected. Investigating the semantic information of the collective data plays an important role in understanding the needs, analyzing preference of the user, even recommending and predicting space and time. Recently, many researchers all over the world have turned their focus on understanding the spatio-temporal semantic data. This paper summarizes the related works regarding the spatio-temporal semantic data. Firstly, according to the tasks, the basic concepts and research frameworks are introduced; then, the works of location semantic understanding, user behavior semantic understanding and event semantic understanding are summarized. Additionally, the application scenarios of recommending and predicting space and time field are described. Finally, the future research directions of spatio-temporal data semantic understanding are discussed. |
| Key words: spatio-temporal data mining semantic understanding semantic profiling user portrait machine learning |
