日本データベース学会の皆様， NTT　CS研　木村です． 下記の通り，ICME2014併設のワークショップを企画致しております． 投稿をご検討いただければ幸いです．よろしくお願い致します． -- Call for Papers 1st International Workshop on Cross-media Analysis from Social Multimedia (CASM), in conjunction with ICME2014, July 14, 2014, Chengdu, China. http://sites.google.com/site/casm2014ws/ http://twitter.com/casm2014/ Submission deadline: March 23, 2014 Author notification: April 9, 2014 Camera-ready submission deadline: April 16 We are entering the age of ubiquitous social multimedia. User-generated multimedia contents such as blogs, short messages, photos and location check-ins become so pervasive that it is now feasible to exploit them as sensors for real-world and web-world objects, events and memes. As such, new algorithms for analyzing unorganized user-generated contents have become one of the most active research areas in several research communities. Also, a recent phenomenon called content curation, such as Storify and Pinterest is now becoming an emerging trend in social media, which helps us to discover, analyze and retrieve underlying contexts of user-generated multimedia contents. However, the ability to deeply analyze and understand those unorganized contents on a single platform still remains inadequate, even with state-of-the-art multimedia processing and machine learning techniques.

From the above background, this workshop focuses on one of emerging but challenging researches topics, cross-media analysis of user-generated and unorganized social multimedia contents on multiple platforms. Those social media contents include rich privileged information such as geo locations, descriptions, user profiles, network and content diffusion that enable us to view and investigate one specific object, event or meme from various aspects. This workshop also serves for recent advances in harvesting and mining multimedia contents along with various types of privileged information.

This workshop is devoted to the publications of high quality papers on innovative technical developments and practical applications related to cross-media analysis from social media, such as: - Real-world event detection, localization, tracking and understanding from social multimedia - Trend detection, diffusion modeling and anomaly detection of social multimedia - Context representation, modeling and understanding of multimedia contents with multiple privileged information from social media - Sentiment analysis, preference analysis, and user profiling from social multimedia - Content analysis, indexing, retrieval and harvesting of social multimedia - Multimedia content browsing, visualization and mining over multiple social media platforms - Mobile cross-media analysis, summarization, visualization and browsing - Transfer learning, semi-supervised learning and domain adaptation across multiple social media resources Authors are invited to submit a paper (two columns, 6 pages maximum) according to the guidelines available on the ICME2014 website. Only electronic submissions will be accepted. This workshop uses double-blind review process for all the regular submissions. Any papers violating double-blind regulations will be immediately rejected without reviews. Simultaneous submission to other conferences in the area of multimedia and related areas is not permitted. Simultaneous submission to journal publications of significantly extended versions of the paper is permitted, as long as the publication date of the journal is after the workshop date. Workshop organizers: Akisato Kimura, NTT Communication Science Laboratories, Japan Winston Hsu, National Taiwan University, Taiwan Shin'ichi Satoh, National Institute of Informatics, Japan Contact organizers at casm2014-org at googlegroups com -- ----- Akisato Kimura, Ph.D Recognition Research Group Media Information Laboratory NTT Communication Science Laboratories E-mail: akisato@ieee.org URL: http://www.brl.ntt.co.jp/people/akisato/ The effort may not be rewarded. But nothing great was ever achieved without effort.