Homepage: http://www.livememories.org

Collective Digital Memories

The ubiquitousness of low-cost means for producing digital photos or videos (digital cameras, cellular phones) together with the explosion in use of social networking sites for sharing such  data and / or one's opinions about the events of the day on the Web (Flickr, Facebook, blogs) are leading to radically  new opportunities for the creation of collective  memory. However, much of the potentiality of this new way of taking advantage of the Web is not exploited by current sites, where such data lie unanalyzed. 

Making Digital Memories Alive through Content Extraction

The aim of LiveMemories, a PAT Grande Progetto in collaboration with  FBK, UniTN (DISI / CIMEC), and University of Southampton, is to take advantage of content extraction techniques to make such shared digital memories alive by highlighting links with other digital memories and by  enabling new presentation methods, as in the following scenario. Luisa Tomasi goes to the Franco Battiato concert in Trento on February 17th and 18th, and takes some pictures. The next day she creates a description of the event on the LiveMemories portal, uploading the images and accompanying them with some text describing her experience. Images and text are analyzed by the LiveMemories platform, that recognizes the event as one listed on www.crushsite.it and  identifies Franco Battiato as one of the individuals stored in its knowledge base (automatically extracted by processing WikiPedia text and text from the local press), someone with his own website. LiveMemories can then offer to Luisa further information about the event.g., it can tell Luisa that the brilliant viola player is called Demetrio Comuzzi, or it may offer to Luisa to visualize Battiato's discography in the form of a chronology to discover when a particular song came out. LiveMemories may also discover that other people with an account on the portal went to that same concert e.g. Mario Boato, who also uploaded his own data. LiveMemories can point this out to Luisa and Mario, who may also discover they share a preference for Battiato's early music.

Background

LiveMemories is the natural extension of several projects at the participating institutions, including

• OntoText at FBK (tcc.itc.it/projects/ontotext/) , a project on content extraction and entity disambiguation that also resulted in the creation of the TextPro NLP suite (textpro.fbk.eu/);
• The ELERFED Johns Hopkins 2007 Workshop  (www.clsp.jhu.edu/ws2007/groups/elerfed/) that produced the intra-document coreference toolkit BART (www.assembla.com/wiki/show/bart-coref);
• The development of the SWeb platform
• TAGORA, a project on collaborative tagging (www.tagora-project.eu/) 
• The CLEANEVAL and WACKY Initiatives for creating large, clean corpora from the Web (wacky.sslmit.unibo.it/doku.php)
• NEEDLE (http://odle.dit.unitn.it:8080/needleNew/)
• RECK (http://cicerone.dit.unitn.it:8180/RECK/).

Web Science 

The University of Southampton is in partnership with MIT in Web Science, a joint  initiative that has a goal to facilitate and produce the fundamental scientific advances necessary to inform the future design and use of the World Wide Web.  LiveMemories will contribute to this effort by evaluating the feasibility and the contribution of  content extraction on the scale of the Web.

Members

Faculty Members

Fausto Giunchiglia	Alessandro Moschitti	Massimo Poesio
Giuseppe Riccardi	Marco Ronchetti

Doctoral Students

S. Ghosh	T. Nguyen Thi	J. Pane

Partners

CIMEC
Department of Sociology and Social Research
FBK-IRST
Università degli Studi di Trento
University of Southampton - School of Electronics and Computer Science