Défi Masses de Données Scientifiques – MASTODONS
Crowdsourcing and human computation multidisciplinary workshop
- Day : 2014/09/15
- Location : Auditorium Marie Curie CNRS, Paris, 75016
- Final program
- Click here to register (registration is compulsory but free)
- Official web page
The crowdsourcing and human computation multidisciplinary workshop is organized by the CNRS MASTODONS* challenge.
Crowdsourcing systems are becoming very popular in a variety of domains and for a wide range of online tasks. Common crowdsourcing scenarios include data gathering (asking volunteers to tag a picture or a video), document editing (as in Wikipedia), opinion solicitation (asking foodies to provide a summary of their experience at a restaurant), and collaborative intelligence (asking residents to match old city maps). Human computation addresses the design or analysis of information processing systems in which human groups participate as computational elements. Both fields are concerned with how people can be involved in solving complex tasks that go beyond the capabilities of current algorithms and with the emergence of collective intelligence. Research in these domains is interdisciplinary and includes efforts in human computer interaction, cognitive psychology, economics, data bases, information retrieval, artificial intelligence, etc. The purpose of this one-day interdisciplinary workshop is to bring together researchers and practitioners from different fields to share their experience and thoughts about the challenges and opportunities of doing (or not !) research on crowdsourcing and human computation.
Confirmed keynote speakers :
Atsuyuki Morishima, Professor at Tsukuba University, http://www.kc.tsukuba.ac.jp/~mori/
CyLog/Crowd4U : Toward an Earth-scale Volunteer Network for Microtask-based Crowdsourcing
Abstract : This talk overviews FusionCOMP project that addresses problems in software engineering for data-centric crowdsourcing applications. The project started in 2009 and have developed a programming language named CyLog and its execution platform Crowd4U. CyLog is a datalog-like language that allows some predicates to be evaluated by humans. Crowd4U has an execution engine for CyLog codes and is being used for public and academic crowdsourcing projects involving more than 20 universities. This talk explains our challenges and some of the results we obtained in the project.
Bio : Atsuyuki Morishima is a professor of research center for knowledge communities, faculty of library, information, and media science, University of Tsukuba, Japan. His research interests include data integration, Web data management, and integration of data-centric human/machine computations. His SIGMOD 2001 paper on XML data publishing was the second runner-up for the best paper award. In 2004, he received the IPSJ best paper award. In 2009, he received Kanbayashi award from DBSJ (the Database Society of Japan). In 2013, he received the best paper award of DBSJ Journal. Currently， he is the chair of SIG-DBS of Information Processing Society of Japan. In the past, he served as a steering committee member of ACM SIGMOD Japan Chapter. He served as program committee members and conference officers of conferences including VLDB, SIGMOD, ICDE, EDBT, CIKM, ECDL, DASFAA, ICADL, XSym, PersDB, and DBCrowd. He received his Ph.D degree in computer science from University of Tsukuba in 1998.
Evelyne Lutton, Senior Research Scientist, INRIA, France, http://evelyne.lutton.free.fr/
Humanized computation and interactive evolution
Abstract : This talk will be focussed on questions related to the combination of autonomous computer calculations with user interactions, often called humanized computation. Classical approaches like interactive evolution consider the human in the loop only as an information provider (like human based evaluation for an optimisation process for instance). Recent developments in this domain progressively start to consider the human user as a full component of the process, for instance providing « human calculated » solutions to the process. Common characteristics with crowdsourcing issues become thus evident, in particular when the system is interacting with many users.
Arnaud Vincent, founder of AVIOMEX SAS, http://www.aviomex.com
Krabott, When the amateurs take over a market trading room
Abstract : With more than 300 « players », the human computation project Krabott, has been running, for 5 years now, an open experimentation on real-time market data. Krabott fosters large-scale human collaboration for the elaboration of systematic trading strategies able to outperform algorithms based on market expert knowledge. Krabott was the subject of a thesis defended in November 2013 at Mines ParisTech « Human computation applied to automated trading »
MASTODONS is a challenge on the exploitation of large scientific data sets, launched by the multidisciplinary mission of CNRS