Special Issue on Deep Learning for Multi-modal Social Media Analysis and Applications
摘要截稿:
全文截稿: 2019-07-15
影响因子: 4.787
期刊难度:
CCF分类: B类
中科院JCR分区:
• 大类 : 计算机科学 - 1区
• 小类 : 计算机:信息系统 - 1区
• 小类 : 图书情报与档案管理 - 1区
Overview
1. Aims and Scopes
Recent years have witnessed the proliferation of social media (e.g., Twitter, Instagram and Foursquare), which greatly facilitates the web users to connect, interact and share information to the others. The large number of user-generated contents (UGCs) on social media have attracted great attentions from various research communities, including the data mining, information retrieval and multimedia analysis. Meanwhile, with the boom of big data, deep learning methods are enabled to achieve compelling success in many research tasks, such as the image classification, sentiment analysis and machine translation. In a sense, the advanced deep neural networks, such as the Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs) or Long-Short Term Memory (LSTM), have enabled us to learn the better representations of the mono-modal data. Nevertheless, the emerging real-world social media data gradually exhibits the complex nature, which usually involves multiple modalities, such as the text, image, video and audio. The complex relation among different modalities and the heterogeneity of the multi-modal data hence poses new challenges for social media analysis and applications with the conventional deep neural networks, including the benchmark dataset construction, multi-modal representation learning, multi-modal semantic modeling, multi-modal data fusion and knowledge discovery from the multi-modal data.
This special issue aims to bring together the innovative researches across the world in this interesting research area. In particular, we expect the novel contributions focus on the following research lines: (1) state-of-the-art techniques for the multi-modal social media analysis; (2) novel applications based on the emerging multi-modal social media data; (3) surveys of recent progress in this research area; and (4) the benchmark dataset construction.
2. Topics of Interest
The list of possible topics includes, but is not limited to:
Deep neural networks for multi-modal social media representation learning
Deep neural networks for multi-modal social media fusion
Deep neural networks for multi-modal social media semantic modeling
Deep neural networks for multi-modal social media summarization
Deep neural networks for knowledge discovery from multi-modal social media
Deep neural networks for cross-modal social media retrieval
Deep neural network for cross-modal social media generation (e.g., image/video captioning)
Deep neural networks for emerging multi-modal applications with social media, such as the fashion analysis and smart healthcare
New dataset and benchmark for multi-modal social media analysis