摘要
随着互联网和社会媒体平台的发展,社会媒体吸引着数以亿计的用户参与其中进行创造和分享信息,产生了海量的文本、图像、音频和视频数据.面对这些数量巨大、异构多源、模态复杂的社会多媒体内容数据,如何对其进行有效的内容理解和知识表示,从而为用户提供更高效、优质的服务,成为实现社会媒体大数据价值的关键.本文对近年来在社会多媒体内容分析、知识提取和表示以及用户建模应用的相关研究展开综述,并针对社会多媒体特征融合、跨模态知识提取与表示,以及基于社会媒体的用户建模相关应用研究三个方面进行详细总结.随后对社会多媒体内容的知识表示和用户建模的研究与应用的发展趋势进行介绍,最后对多媒体知识表示与用户建模研究进行了总结和展望.
With rapid development of the Internet and social media platforms,hundreds of millions of users have created and shared information as a large quantities of text,image,audio,and video data.Given this large volume of heterogeneous,multi-source,and complex social multimedia content,the challenge of providing effective content understanding and knowledge representation to users with more efficient and high-quality service has become the key to realizing the value of big social media data.This paper reviews research on social multimedia content analysis,knowledge extraction and representation,and on user modeling applications in recent years.In addition,research on social multi-modal feature fusion,cross-modal knowledge extraction and representation,and user modeling based on social media has been presented in detail.We then discuss development trends in social multimedia content knowledge representation,and review research into and application of user modeling.Finally,we summarize likely research directions for multimedia knowledge representation and user modeling.
作者
徐常胜
黄晓雯
钱胜胜
方全
XU Changsheng;HUANG Xiaowen;QIAN Shengsheng;FANG Quan(National Lab of Pattern Recognition,Institute of Automation,CAS,Beijing 100190;University of Chinese Academy of Sciences,Beijing 100190)
出处
《南京信息工程大学学报(自然科学版)》
CAS
2020年第1期31-44,I0008-I0013,共20页
Journal of Nanjing University of Information Science & Technology(Natural Science Edition)
基金
国家自然科学基金(61432019,61720106006,61572503,61802405,61872424,61702509,61832002)
国家重点研发计划(2017YFB1002804)。
关键词
社会多媒体
多模态
知识表示
用户建模
social multimedia
multi-modality
knowledge representation
user modeling