1,721,045 research outputs found
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Compromis entre confidentialité et utilité dans la prise de décision séquentielle dans l’incertain
No full textLes thèmes abordés dans cette thèse visent à caractériser les compromis à réaliser entre confidentialité et utilité dans la prise de décision séquentielle dans l'incertain. Le principal cadre adopté pour définir la confidentialité est la protection différentielle, et le principal cadre d'utilité est le problème de bandit stochastique à plusieurs bras. Tout d'abord, nous proposons différentes définitions qui étendent la définition de confidentialité à l'environnement des bandits à plusieurs bras.Ensuite, nous quantifions la difficulté des bandits avec protection différentielle en prouvant des bornes inférieures sur la performance des algorithmes de bandits confidentielles. Ces bornes suggèrent l'existence de deux régimes de difficulté en fonction du budget de confidentialité et des distributions de récompenses.Nous proposons également un plan générique pour concevoir des versions confidentielles quasi-optimales des algorithmes de bandits.Nous instancions ce schéma directeur pour concevoir des versions confidentielles de différents algorithmes de bandits dans différents contextes: bandits à bras finis, linéaires et contextuels avec le regret comme mesure d'utilité, et bandits à bras finis avec la complexité d'échantillonnage comme mesure d'utilité.L'analyse théorique et expérimentale des algorithmes proposés valide aussi l'existence de deux régimes de difficulté en fonction du budget de confidentialité.Dans la deuxième partie de cette thèse, nous passons des défenses de la confidentialité aux attaques. Plus précisément, nous étudions les attaques par inférence d'appartenance où un adversaire cherche à savoir si un point cible a été inclus ou pas dans l'ensemble de données d'entrée d'un algorithme. Nous définissons la fuite d'information sur un point comme l'avantage de l'adversaire optimal essayant de déduire l'appartenance de ce point.Nous quantifions ensuite cette fuite d'information pour la moyenne empirique et d'autres variantes en termes de la distance de Mahalanobis entre le point cible et la distribution génératrice des données.Notre analyse asymptotique repose sur une nouvelle technique de preuve qui combine une expansion de Edgeworth du test de vraisemblance et un théorème central limite de Lindeberg-Feller.Notre analyse montre que le test de vraisemblance pour la moyenne empirique est une attaque par produit scalaire mais corrigé pour la géométrie des données en utilisant l'inverse de la matrice de covariance.Enfin, comme conséquences de notre analyse, nous proposons un nouveau score de covariance et une nouvelle stratégie de sélection des points cible pour l'audit des algorithmes de descente de gradient dans le cadre de l'apprentissage fédéré en white-box.The topics addressed in this thesis aim to characterise the privacy-utility trade-offs in sequential decision-making under uncertainty. The main privacy framework adopted is Differential Privacy (DP), and the main setting for studying utility is the stochastic Multi-Armed Bandit (MAB) problem. First, we propose different definitions that extend DP to the setting of multi-armed bandits. Then, we quantify the hardness of private bandits by proving lower bounds on the performance of bandit algorithms verifying the DP constraint. These bounds suggest the existence of two hardness regimes depending on the privacy budget and the reward distributions. We further propose a generic blueprint to design near-optimal DP extensions of bandit algorithms. We instantiate the blueprint to design DP versions of different bandit algorithms under different settings: finite-armed, linear and contextual bandits under regret as a utility measure, and finite-armed bandits under sample complexity of identifying the optimal arm as a utility measure. The theoretical and experimental analysis of the proposed algorithms furthermore validates the existence of two hardness regimes depending on the privacy budget.In the second part of this thesis, we shift the view from privacy defences to attacks. Specifically, we study fixed-target Membership Inference (MI) attacks, where an adversary aims to infer whether a fixed target point was included or not in the input dataset of an algorithm. We define the target-dependent leakage of a datapoint as the advantage of the optimal adversary trying to infer the membership of that datapoint. Then, we quantify both the target-dependent leakage and the trade-off functions for the empirical mean and variants of interest in terms of the Mahalanobis distance between the target point and the data-generating distribution. Our asymptotic analysis builds on a novel proof technique that combines an Edgeworth expansion of the Likelihood Ratio (LR) test and a Lindeberg-Feller central limit theorem. Our analysis shows that the LR test for the empirical mean is a scalar product attack but corrected for the geometry of the data using the inverse of the covariance matrix. Finally, as by-products of our analysis, we propose a new covariance score and a new canary selection strategy for auditing gradient descent algorithms in the white-box federated learning setting
Interactive and Concentrated Differential Privacy for Bandits
No full textInternational audienceBandits play a crucial role in interactive learning schemes and modern recommender systems. However, these systems often rely on sensitive user data, making privacy a critical concern. This paper investigates privacy in bandits with a trusted centralized decision-maker through the lens of interactive Differential Privacy (DP). While bandits under pure -global DP have been well-studied, we contribute to the understanding of bandits under zero Concentrated DP (zCDP). We provide minimax and problem-dependent lower bounds on regret for finite-armed and linear bandits, which quantify the cost of -global zCDP in these settings. These lower bounds reveal two hardness regimes based on the privacy budget and suggest that -global zCDP incurs less regret than pure -global DP. We propose two -global zCDP bandit algorithms, AdaC-UCB and AdaC-GOPE, for finite-armed and linear bandits respectively. Both algorithms use a common recipe of Gaussian mechanism and adaptive episodes. We analyze the regret of these algorithms to show that AdaC-UCB achieves the problem-dependent regret lower bound up to multiplicative constants, while AdaC-GOPE achieves the minimax regret lower bound up to poly-logarithmic factors. Finally, we provide experimental validation of our theoretical results under different settings
Concentrated Differential Privacy for Bandits
Full text linkBandits serve as the theoretical foundation of sequential learning and an
algorithmic foundation of modern recommender systems. However, recommender
systems often rely on user-sensitive data, making privacy a critical concern.
This paper contributes to the understanding of Differential Privacy (DP) in
bandits with a trusted centralised decision-maker, and especially the
implications of ensuring zero Concentrated Differential Privacy (zCDP). First,
we formalise and compare different adaptations of DP to bandits, depending on
the considered input and the interaction protocol. Then, we propose three
private algorithms, namely AdaC-UCB, AdaC-GOPE and AdaC-OFUL, for three bandit
settings, namely finite-armed bandits, linear bandits, and linear contextual
bandits. The three algorithms share a generic algorithmic blueprint, i.e. the
Gaussian mechanism and adaptive episodes, to ensure a good privacy-utility
trade-off. We analyse and upper bound the regret of these three algorithms. Our
analysis shows that in all of these settings, the prices of imposing zCDP are
(asymptotically) negligible in comparison with the regrets incurred oblivious
to privacy. Next, we complement our regret upper bounds with the first minimax
lower bounds on the regret of bandits with zCDP. To prove the lower bounds, we
elaborate a new proof technique based on couplings and optimal transport. We
conclude by experimentally validating our theoretical results for the three
different settings of bandits.Comment: Appears in IEEE SaTML 202
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
The Fair Game: Auditing & debiasing AI algorithms over time
No full textInternational audienceAbstract An emerging field of AI, namely Fair Machine Learning (ML), aims to quantify different types of bias (also known as unfairness) exhibited in the predictions of ML algorithms, and to design new algorithms to mitigate them. Often, the definitions of bias used in the literature are observational, i.e. they use the input and output of a pre-trained algorithm to quantify a bias under concern. In reality, these definitions are often conflicting in nature and can only be deployed if either the ground truth is known or only in retrospect after deploying the algorithm. Thus, there is a gap between what we want Fair ML to achieve and what it does in a dynamic social environment. Hence, we propose an alternative dynamic mechanism, “Fair Game”, to assure fairness in the predictions of an ML algorithm and to adapt its predictions as the society interacts with the algorithm over time. “Fair Game” puts together an Auditor and a Debiasing algorithm in a loop around an ML algorithm. The “Fair Game” puts these two components in a loop by leveraging Reinforcement Learning (RL). RL algorithms interact with an environment to take decisions, which yields new observations (also known as data/feedback) from the environment and in turn, adapts future decisions. RL is already used in algorithms with pre-fixed long-term fairness goals. “Fair Game” provides a unique framework where the fairness goals can be adapted over time by only modifying the auditor and the different biases it quantifies. Thus, “Fair Game” aims to simulate the evolution of ethical and legal frameworks in the society by creating an auditor which sends feedback to a debiasing algorithm deployed around an ML system. This allows us to develop a flexible and adaptive-over-time framework to build Fair ML systems pre- and post-deployment
- …
