HAL - Lille 3
Not a member yet
5967 research outputs found
Sort by
Une nouvelle approche linguistique pour évaluer l'opinion des usagers dans les réseaux sociaux
No full textInternational audienceThis article describes an automated technique that allows to differentiate texts expressing a positive or a negative opinion. The basic principle is based on the observation that positive texts are statistically shorter than negative ones. From this observation of the psycholinguistic human behavior, we derive a heuristic that is employed to generate connoted lexicons with a low level of prior knowledge. The lexicon is then used to compute the level of opinion of an unknown text. Our primary motivation is to reduce the need of the human implication (domain and language) in the generation of the lexicon in order to have a process with the highest possible autonomy. The resulting adaptability would represent an advantage with free or approximate expression commonly found in social networks environment
Space-time Histograms And Their Application To Person Re-identification In TV Shows
No full textInternational audienceThe annotation of video streams by automatic content analysis is a growing field of research. The possibility of recognising persons appearing in TV shows allows to automatically structure ever-growing video archives. We propose a new descriptor to re-identify persons featured in videos, that is to say, to spot all occurrences of persons throughout a video. Our approach is dynamic as it benefits from motion information contained in videos, whereas the static approaches are solely based on still images. We extract persontracks from videos and match them using a new descriptor and its associated similarity measure: the space-time histogram. The originality of our approach is the integration of temporal data into the descriptor. Experiments show that it provides a better estimation of the similarity between persontracks. Our contribution has been evaluated using a corpus of real life french TV shows broadcasted on BFMTV and LCP TV channels and on some annotated episodes from “Buffy: the Vampire Slayer”. Experimental results show that our approach significantly improves the precision of the re-identification process thanks to the use of the temporal dimension
Fault detection and exclusion method for a tightly coupled localization system
No full textInternational audienc
"Structure, standards and Stoic moral progress in De Finibus 4"
No full textInternational audienceThis paper aims at showing the implicit structure and criteria of the refutation of Stoicim in Cicero's De finibus 4, accounting for a number of puzziing features of the text, and for its coherence with books 3 and 5 of De finibus
A First Analysis of String APIs: the Case of Pharo
No full textInternational audienceMost programming languages natively provide an abstraction of character strings. However, it is difficult to assess the design or the API of a string library. There is no comprehensive analysis of the needed operations and their different variations. There are no real guidelines about the different forces in presence and how they structure the design space of string manipulation. In this article, we harvest and structure a set of criteria to describe a string API. We propose an analysis of the Pharo 4 String library as a first experience on the topic
Truthful Learning Mechanisms for Multi–Slot Sponsored Search Auctions with Externalities
No full textInternational audienceSponsored Search Auctions (SSAs) constitute one of the most successful applications of microeconomic mechanisms. In mechanism design, auctions are usually designed to incentivize advertisers to bid their truthful valuations and, at the same time, to guarantee both the advertisers and the auctioneer a non–negative utility. Nonetheless, in sponsored search auctions, the Click–Through–Rates (CTRs) of the advertisers are often unknown to the auctioneer and thus standard truthful mechanisms cannot be directly applied and must be paired with an effective learning algorithm for the estimation of the CTRs. This introduces the critical problem of designing a learning mechanism able to estimate the CTRs at the same time as implementing a truthful mechanism with a revenue loss as small as possible compared to the mechanism that can exploit the true CTRs. Previous work showed that, when dominant–strategy truthfulness is adopted, in single–slot auctions the problem can be solved using suitable exploration–exploitation mechanisms able to achieve a cumulative regret (on the auctioneer's revenue) of order , where is the number of times the auction is repeated. It is also known that, when truthfulness in expectation is adopted, a cumulative regret (over the social welfare) of order can be obtained. In this paper, we extend the results available in the literature to the more realistic case of multi–slot auctions. In this case, a model of the user is needed to characterize how the CTR of an ad changes as its position in the allocation changes. In particular, we adopt the cascade model, one of the most popular models for sponsored search auctions, and we prove a number of novel upper bounds and lower bounds on both auctioneer’s revenue loss and social welfare w.r.t. to the Vickrey–Clarke–Groves (VCG) auction. Furthermore, we report numerical simulations investigating the accuracy of the bounds in predicting the dependency of the regret on the auction parameters
Imitation Learning Applied to Embodied Conversational Agents
No full textInternational audienceEmbodied Conversational Agents (ECAs) are emerging as a key component to allow human interact with machines. Applications are numerous and ECAs can reduce the aversion to interact with a machine by providing user-friendly interfaces. Yet, ECAs are still unable to produce social signals appropriately during their interaction with humans, which tends to make the interaction less instinctive. Especially, very little attention has been paid to the use of laughter in human-avatar interactions despite the crucial role played by laughter in human-human interaction. In this paper, methods for predicting when and how to laugh during an interaction for an ECA are proposed. Different Imitation Learning (also known as Apprenticeship Learning) algorithms are used in this purpose and a regularized classification algorithm is shown to produce good behavior on real data
