A new blog article has been published on the PuppyIR blog. The blog can be found here.

In this invited talk, Azzopardi talked about the importance of addressing and supporting the information needs of children.The early adoption of future technologies by children create a new digital world which presents numerous challenges to IR research, in particularly in how children learn to search, and search to learn.

PuppyIR will organize an Exploitation Workshop the next 13^th of March 2012. This event is mostly oriented to public sector, health and learning areas of interest. The venue will be Museon, Stadhouderslaan 37, 2517 HV Den Haag, Netherlands.

UGLW (Ric Glassey and Leif Azzopardi) prepared teaching material for a one-day teaching session on the theme of “Helping Children Find Information” and learning about the PuppyIR Project.
Target audience: MSc and early-PhD students.
Website
Slides

Carsten Eickhoff (TUD), Andreas Lingnau (UoS)
Slides: [NEM2011]

Slate Magazine spent an article on the techniques PuppyIR introduces to help make the internet a safer place for children and to support them with age-sensitive search tools. It mentions the Youtube classification algorithm¹ developed by Carsten Eickhoff and Arjen P. de Vries (TUDelft) and the AgeRank technique² by Karl Gyllstrom and Marie-Francine Moens (KULeuven). The full Slate article can be found here

1. Carsten Eickhoff, Arjen P. de Vries, Identifying Suitable YouTube Videos for Children, Proceedings of the 3rd Networked & Electronic Media Summit (NEM),  
2. Karl Gyllstrom, Marie-Francine Moens, L., Wisdom of the Ages: Toward Delivering the Children’s Web with the Link-based AgeRank Algorithm, Proceedings of the International Conference in Information and Knowledge Management (CIKM).  

To show the possibilities of the PuppyIR framework two demonstrators are being constructed, one for Emma Children’s Hospital and one for Museon. The first results of this latter demonstrator are already available to the visitor of the museum, where PuppyIR technology is used in an educational, collaborative quest through Museon’s permanent exhibition.

At the entrance of the museum a multitouch table is located that teams of three or four players can use to personalize their quest: via collaborative interaction they determine their quest’s contents. Based on the interaction at the multitouch each team member get their own assignments, but since all assignments are related to the same subjects they are complementary. The result is that during the quest co-operation and exchange of information are stimulated, which enhances the educational value of the game. During the quest the team collects items for the end game back at the multitouch table, that aims at applying the information that is obtained during the quest: the team has to match the items that they collected with relevant concepts.

Currently the PuppyIR team is working on an advanced version of the game, turning it into a genuine full circle experience, starting in the museum and ending at home. After the actual quest the players will be able to dig deeper into contents of the exhibits they visited and to compose a personal website. Back home, by showing the fiducial they used in the museum to a webcam, a search is launched that brings forward information that is related to the contents of their quest, coming from the museum’s website but also from elsewhere on the web.

This video shows the new PuppyIR demonstrator that is still under construction.

Carsten Eickhoff (Delft University of Technology) and Arjen de Vries (CWI and Delft University of Technology) have won the Microsoft Bing Most Innovative Award, awarded at the CSDM workshop February 2011 in Hong Kong. Their award winning paper “How Crowdsourcable is Your Task?” aims to identify measures to make a crowdsource task more resistant to fraudulent attempts (awarded for creativity, originality and potential impact).

Crowdsourcing is a relatively new concept. Briefly, it means that a task unexecutable by a computer is being outsourced to human volunteers, sometimes against a small fee. An example is judging whether a website is suitable for children. This way, in a short amount of time large databases can be created that are useful to all kinds of research. Crowdsourcing is sensitive to people who are after the fee and consequently generate random answers. This happens frequently and it makes it hard to interpret data obtained by crowdsourcing the right way. Eickhoff and de Vries have proposed measures to help banish these kind of practices.

The Microsoft Bing Innovative Award is recognized to the most novel and innovative paper on crowdsourcing. Selection criteria are creativity, originality and potential impact of the described proposal.

The collection of labeled data for the web page classification research using crowdsourcing shows that PuppyIR is well aware of the necessity to get high accuracy levels in data labels.

Proposed and organized by Pavel Serdyukov, Djoerd Hiemstra and Ian Ruthven (ACM SIGIR Conference on Research and Development in Information Retrieval, Geneva, Switzerland, 2010)

• PuppyIR presentation
• Chorus Website