Mar 22

Congrats André and Luca!

We are proud that, on February 27th, Luca M. Aiello and André Panisson received their Ph.D. in Computer Science from the Doctoral School of Sciences and Innovative Technologies, Computer Science Programme, University of Turin.

Now they moved elsewhere to spread (and to keep learning) Science and Technology across Europe 🙂

Currently, André is a Junior Researcher at ISI Foundation in Turin and Luca is a post-doc at Yahoo! Research in Barcelona. This is what we call an outstanding debut!

Good luck guys: we’ll keep following you!

Dec 08

DataInterfaces.org cited by “La Stampa” (December 06, 2011)

Italian newspaper “La Stampa” cited datainterfaces project, that involves Politecnico di Milano, ISI, and our group, in an article commenting the “Big Data in a Living Web” conference held in Turin, 2011/12/06.

Moreover, during the conference, the video by our André Panisson on retweets of Twitter messages during the Egyptian revolution in February 11, 2011, has been presented as an example of innovative complex data visualization strategy.

Nov 15

ARCS Group at TOSM 2011

ARCS Group is present at TOSM (Torino Software and Systems Meeting , 16-18 November). We will show latest advances on MobHinter, a recommender system for mobile devices proposed originally in 2008 and recently developed as part of AS4.MOBI project. MobHinter is designed for spreading of information on a geographic proximity basis, and it allows for timely advertising and notification of contex-aware events.

Nov 02

New paper in ACM Transaction on the Web (TWEB)

The paper “Friendship prediction and homophily in social media” has been accepted for publication on the ACM Transaction on the Web (TWEB) journal.

Abstract

Web 2.0 applications have attracted considerable attention because their open-ended nature allows users to create lightweight semantic scaffolding to organize and share content. To date, the interplay of the social and topical components of social media has been only partially explored. Here we study the presence of homophily in three systems that combine tagging of social media with online social networks. We find a substantial level of topical similarity among users who lie close to each other in the social network. We introduce a null model that preserves user activity while removing local correlations, allowing us to disentangle the actual local similarity between users from statistical effects due to the assortative mixing of user activity and centrality in the social network. This analysis suggests that users with similar interests are more likely to be friends, and therefore topical similarity measures among users based solely on their annotation metadata should be predictive of social links. We test this hypothesis on several datasets, confirming that social networks constructed from topical similarity capture actual friendship accurately.

Oct 10

INCA project Approved!

The INCA (Interactive Content Architecture) project was approved as part of the Poles of Innovation “Polo della Creatività Digitale e Multimedialità” and “Polo ICT“. Submitted by ARCS Group in collaboration with euriX s.r.l., Cedeo s.a.s., synArea and CSP, the INCA project aims to develop all necessary technologies to support the life cycle of transmedia and 3D content (named INCA content) in order to establish an international standard of reference, and then offer the ability to:

 

  1. Authoring with management/protection functionality  of intellectual property and INCA content “user-generated” (tag) description
  2. Efficient storage of INCA content for further search and streaming
  3. Navigational search of INCA content based on user knowledge and its social network
  4. Streaming broadcast of INCA content for a variety of deployment environments such as web, IPTV, mobile and broadcast
  5. Fruition of INCA content

Oct 05

ARCS group is on line!

 

Hello World! We are on line now, with a brand new site! Follow us if you want know what we are doing and where we are going. Of course, if you like our posts, feel free to share with your friends what you find here on your preferred social media.

Oct 01

Wall Street Journal on Truthy

New Wall Street Journal article with a video on the Truthy Project at Indiana University, featuring also the movie generation tool.  Congratulations to all the Truthy team and to Luca Maria Aiello that contributed the tool!

Aug 27

New paper in ACM Transactions

New paper “Contextual tag inference” accepted on TOMMCAP, ACM Transactions on Multimedia Computing, Communications and Applications.

Abstract:”This paper examines the use of two kinds of context to improve the results of content-based music taggers: the relationships between tags and between the clips of songs that are tagged. We show that users agree more on tags applied to clips temporally “closer” to one another; that conditional restricted Boltzmann machine models of tags can more accurately predict related tags when they take context into account; and that when training data is “smoothed” using context, support vector machines can better rank these clips according to the original, unsmoothed tags and do this more accurately than three standard multi-label classifiers. “

Aug 20

Paper presentation on EC-Web 2011

The paper “On Collaborative Filtering Techniques for Live TV and Radio Discovery and Recommendation” will be presented in the 12th International Conference on Electronic Commerce and Web Technologies (EC-Web 2011), at 31 August in the Session 5: Collaborative Filtering and Preference Learning. The EC-Web 2011 Conference will happen within the 22nd DEXA Conference in Toulouse from August 29 to September 2.

Aug 19

Final publication in Springer P2P

The paper “An identity-based approach to secure P2P applications with Likir” has been officially published in Springer Peer-to-Peer networking and Applications Journal.

Abstract:”Structured overlay networks are highly susceptible to attacks aimed at subverting their structure or functionalities. Although many secure architectural design proposals have been presented in the past, a widely accepted and comprehensive solution is lacking. Likir (Layered Identity-based Kademlia-like InfRastructure) is our proposed solution for implementing a secure Peer-to-Peer (P2P) network based on a Distributed Hash Table (DHT). Our purpose is to focus on three main goals: (1) providing security services and a secure overlay infrastructure against the vast majority of security threats on P2P systems, (2) dynamically creating a bridge between randomly generated peer identifiers and user identities, and (3) supplying the developer with a middleware API that can easily deal with peers’ identities. Placing the emphasis on user identity results in a highly secure distributed framework which is very fitting for privacy-aware and efficient implementation of identity-based applications like social networking applications. Detailed security analysis and performance evaluation are provided. Moreover, an implementation of Likir is introduced and a case study is presented in order to show its practical use in a real-life example.”