XRCE organises public scientific seminars on a regular basis which you are welcome to attend. These seminars are an occasion to exchange with researchers from various backgrounds and to broaden scientific expertise. You can subscribe to our seminar RSS Feed for dates, speakers and topics.
In human relationships, language, whether written or spoken, is an inescapable barrier in conveying Expertise or Know-how. Using Artificial Intelligence technology, Yseop generates intelligent, grammatically correct and coherent text in French or English language at the speed of thousands of pages per second. This patented solution constitutes a true driving force between the Know-how of the company's experts or gurus and their operational customer-facing teams in the field.Slides (618.95 kB)
Principal component analysis (PCA) is a classical data analysis technique that finds linear transformations of data that retain maximal amount of variance. We study a case where some of the data values are missing, and show that this problem has many features which are usually associated with nonlinear models, such as overfitting and bad locally optimal solutions. Probabilistic formulation of PCA provides a good foundation for handling missing values, and we introduce formulas for doing that. In case of high dimensional and very sparse data, overfitting becomes a severe problem and traditional algorithms for PCA are very slow. We introduce a novel fast algorithm and extend it to variational Bayesian learning. The scalability of the proposed algorithm is demonstrated by applying it to the Netflix problem.Slides (5.56 MB)
We propose D-STAG, a new formalism for the automatic analysis of the discourse structure of texts. The analyses computed by D-STAG are hierarchical discourse structures annotated with discourse relations, that are compatible with discourse structures computed in Structured Discourse Representation Theory. The discourse analysis extends the sentential analysis, without modifying it, which simplifies the realization of the system. Its current implementation will be discussed.Slides (380.50 kB)
( web e-mail ), Researcher at IDIAP research institute, Martigny, Switzerland will give a talk:
Social Computers for the Social Animal: State-of-the-art and Future Perspectives of Social Signal Processing
Following Aristotle, "Man is by nature a social animal; an individual who is unsocial naturally and not accidentally is either beneath our notice or more than human." This is more than an abstract philosophical statement if, twenty five centuries after the great Greek philosopher, human scientists, biologists and neurologists still investigate how natural evolution has made of humans the perfect machines for social interaction: appropriate brain structures (the mirror neurons) have the explicit goal of making us feel like the people around us, our ears are adapted to the voice of other human beings more than to any other sound, our faces have so many muscles that we can show to others the subtlest changes of our feelings and emotions.
Social Signal Processing (SSP) is the new, emerging, domain that aims at making computers as social as human beings through modeling, analysis and synthesis of nonverbal behavior in social interactions. Social signals are complex aggregates of behavioral cues accounting for our attitude towards others and social situations. They include attention, empathy, politeness, flirting, (dis)agreement, etc., and are conveyed through a multiplicity of cues including posture, facial expression, voice quality, gestures, etc. This talk presents the general concepts underlying SSP, and provides a view of current state-of-the-art. Furthermore, it outlines the most important challenges that the domain has to address to reach its full maturity.Slides (562.01 kB)
( web e-mail ), Assistant Professor at Northwestern University, Chicago, Ill, U.S.A. will give a talk:
Optimal Mechanism Design; from the 2007 Nobel Prize in Economics to the foundation of internet markets
As the Internet has developed to become the most important single arena for resource sharing among parties with diverse and selfish interests, traditional algorithmic and distributed systems approaches are insufficient. To prevent undesirable Internet phenomena such as spam in email systems, bid-sniping in eBay's auction marketplace, free-loading in file-sharing networks, and click-fraud in Internet advertising; game-theoretic and economic considerations from auction theory must be applied. Algorithmic Mechanism Design merges the computational considerations of algorithm and system design with the considerations of the Economics subfield of Mechanism Design. The 2007 Nobel Prize in Economics was awarded to Hurwicz, Maskin, and Myerson, for their foundational work in Mechanism Design recognizing it as fundamental to the field of Economics. Likewise, algorithmic mechanism design is fundamental to the study of computer systems and networks.
This talk provides an introductory sample of topics in the interface of theoretical computer science and the economics subfield of mechanism design.Slides (782.15 kB)