E for Events - Harnessing the Disagreement with Lora Aroyo @ Tagasauris

Session-Level: Intermediate-Advanced
Session-Type: Technology-NLP-New Ideas-Tagging-Media-Linked Data

Date: October 18, 2012
Location: Tagasauris, 175 Varick Street , 10014, New York, NY

Meetup: meetup.com

E for Events - Harnessing the Disagreement with Lora Aroyo

At this session we will take a closer look at events-driven information extraction, data presentation, collection management and metadata enrichment with Lora Aroyo. An ever increasing amount of digital content is unleashed on the Web on a daily basis. This introduces a number of challenges for data managers and end users, e.g. the existing collection metadata process is not sufficient and often not suitable to support the range of user needs and interactions. Similarly collection vocabularies miss links to shared knowledge and the perspective of end users. In this context, we observe a shift from well-curated and closed environments, e.g. museum exhibitions, guide tours, where users were exposed only to a pre-selected set of objects, carefully arranged in stories and narratives. Nowadays, the users have access to an endless pool of information, but the meaning of the individual objects is lost, because of often incomplete descriptions, lack of links between the objects and different collections.

Lora Aroyo is an IBM faculty award winner 2012 for Computer Science. She is an Associate professor for Web and Media at the VU Amsterdam and Scientific coordinator of EU Projects such as NoTube, for integration of Web and TV data with the help of semantics and CHIP, for Cultural Heritage Information Personalization. She is currently on at IBM Research. Her research focuses on using semantic web technologies for modeling user interests and context and applying them in recommendation systems and personalized access to online cultural heritage collections, multimedia archives and interactive TV. She has coordinated the CHIP project on Cultural Heritage Information Personalization and the NoTube project on the integration of Web & TV with the help of semantics. She has co-organized numerous workshops on personalized access to cultural heritage, e-learning, interactive television, visual interfaces to the social and semantic web (e.g. PATCH, FutureTV, PersWeb, VISSW and DeRIVE). Lora is actively involved in the Semantic Web community, i.e. program co-chair for ESWC2009 and ISWC2011 and conference chair for ESWC2010. She is also actively involved in the Personalization and User modeling community as vice-president of UM Inc. and a member of the editorial board for the UMUAI journal.

For more info: http://lora-aroyo.org
For slides: http://www.slideshare.net/laroyo
Follow me on twitter: @laroyo

Tagasauris with Todd Carter

Tagasauris the smart way to tag. Tagasauris is a start up based in New York City that develops innovative services and products .

Todd Carter is the CEO and Co-founder of Tagasauris, a meta-data curation platform that makes your content smarter! We combine crowdsourcing, machine learning and semantic technologies that increases the lifetime value of digital media by making it more discoverable, connected and engaging. Todd is widely respected as a creative visionary and leader in the digital asset, photography and linked open data community, with over 20 years experience working with photo archives, libraries, museums and information technology systems. Tagasauris has been featured in The New York Times, Wired, Business Week, The Economist and others. The National Endowment for the Humanities awarded Tagasauris and The Museum of the City of New York a grant to annotate the museum's archive. Tagasauris is launching it's first photo tagging app that promises to be a game-changer in the media and entertainment industry. Tagasauris was founded in December 2010 with headquarters in New York City.

Dealing with raw data at scale presents present a grand challenge. This is particularly true for visual media where the the large disparity between descriptions of multimedia content that can be computed automatically and the richness and subjectivity of semantics used to find, organize and share visual media present what is called the “semantic gap”. One way to the address this gap would be to relabel raw data. Use experts. Describe the features of visual media with information about their content. But this is impossibly slow and prohibitively expensive! There has to be a better way. Tagasauris has built a large-scale human computation engine that addresses the grand challenge posed by the semantic gap. The tagasauris process automates the discovery and generation of semantically linked metadata and enables us to quickly and cost effectively build a new, interlinked data layer of context for multimedia on the web that "fills in" the semantic gap. The power of this layer will come from the quantity and quality of links between media objects on the visual web and their intersection other objects like events that can be detected and modeled from within the interest and social graphs. We’re interested in discussion and exploration of uses cases that exploit the power of a personalized and semantically interlinked visual web. For more information please visit www.tagasauris.com or email: info@tagasauris.com