arXiv subjects

Suman Banerjee

Publications and source records attributed to Suman Banerjee.

At least 91 records · Page 5Linked to original sources

arXiv2022

Network-Side Digital Contact Tracing on a Large University Campus

We describe a study conducted at a large public university campus in the United States which shows the efficacy of network log information for digital contact tracing and prediction of COVID-19 cases. Over the period of January 18, 2021 to May 7, 2021, more than 216 million client-access-point associations were logged across more than 11,000 wireless access points (APs). The association information was used to find potential contacts for approximately 30,000 individuals. Contacts are determined using an AP colocation algorithm, which supposes contact when two individuals connect to the same WiFi AP at approximately the same time. The approach was validated with a truth set of 350 positive COVID-19 cases inferred from the network log data by observing associations with APs in isolation residence halls reserved for individuals with a confirmed (clinical) positive COVID-19 test result. The network log data and AP-colocation have a predictive value of greater than 10%; more precisely, the contacts of an individual with a confirmed positive COVID-19 test have greater than a 10\% chance of testing positive in the following 7 days (compared with a 0.79% chance when chosen at random, a relative risk ratio of 12.6). Moreover, a cumulative exposure score is computed to account for exposure to multiple individuals that test positive. Over the duration of the study, the cumulative exposure score predicts positive cases with a true positive rate of 16.5% and missed detection rate of 79% at a specified operating point.

cs.NI↗

arXiv2021

Structural Parameterizations of Budgeted Graph Coloring

We introduce a variant of the graph coloring problem, which we denote as {\sc Budgeted Coloring Problem} (\bcp). Given a graph $G$, an integer $c$ and an ordered list of integers $\{b_1, b_2, \ldots, b_c\}$, \bcp asks whether there exists a proper coloring of $G$ where the $i$-th color is used to color at most $b_i$ many vertices. This problem generalizes two well-studied graph coloring problems, {\sc Bounded Coloring Problem} (\bocp) and {\sc Equitable Coloring Problem} (\ecp) and as in the case of other coloring problems, it is \nph even for constant values of $c$. So we study \bcp under the paradigm of parameterized complexity. \begin{itemize} \item We show that \bcp is \fpt (fixed-parameter tractable) parameterized by the vertex cover size. This generalizes a similar result for \ecp and immediately extends to the \bocp, which was earlier not known. \item We show that \bcp is polynomial time solvable for cluster graphs generalizing a similar result for \ecp. However, we show that \bcp is \fpt, but unlikely to have polynomial kernel, when parameterized by the deletion distance to clique, contrasting the linear kernel for \ecp for the same parameter. \item While the \bocp is known to be polynomial time solvable on split graphs, we show that \bcp is \nph on split graphs. As \bocp is hard on bipartite graphs when $c>3$, the result follows for \bcp as well. We provide a dichotomy result by showing that \bcp is polynomial time solvable on bipartite graphs when $c=2$. We also show that \bcp is \nph on co-cluster graphs, contrasting the polynomial time algorithm for \ecp and \bocp. \end{itemize} Finally we present an $\mathcal{O}^*(2^{|V(G)|})$ algorithm for the \bcp, generalizing the known algorithm with a similar bound for the standard chromatic number.

cs.DS↗

arXiv2021

A General Framework For Detecting Anomalous Inputs to DNN Classifiers

Detecting anomalous inputs, such as adversarial and out-of-distribution (OOD) inputs, is critical for classifiers (including deep neural networks or DNNs) deployed in real-world applications. While prior works have proposed various methods to detect such anomalous samples using information from the internal layer representations of a DNN, there is a lack of consensus on a principled approach for the different components of such a detection method. As a result, often heuristic and one-off methods are applied for different aspects of this problem. We propose an unsupervised anomaly detection framework based on the internal DNN layer representations in the form of a meta-algorithm with configurable components. We proceed to propose specific instantiations for each component of the meta-algorithm based on ideas grounded in statistical testing and anomaly detection. We evaluate the proposed methods on well-known image classification datasets with strong adversarial attacks and OOD inputs, including an adaptive attack that uses the internal layer representations of the DNN (often not considered in prior work). Comparisons with five recently-proposed competing detection methods demonstrates the effectiveness of our method in detecting adversarial and OOD inputs.

cs.LG↗

arXiv2021

A Survey on Mining and Analysis of Uncertain Graphs

\emph{Uncertain Graph} (also known as \emph{Probabilistic Graph}) is a generic model to represent many real\mbox{-}world networks from social to biological. In recent times analysis and mining of uncertain graphs have drawn significant attention from the researchers of the data management community. Several noble problems have been introduced and efficient methodologies have been developed to solve those problems. Hence, there is a need to summarize the existing results on this topic in a self\mbox{-}organized way. In this paper, we present a comprehensive survey on uncertain graph mining focusing on mainly three aspects: (i) different problems studied, (ii) computational challenges for solving those problems, and (iii) proposed methodologies. Finally, we list out important future research directions.

cs.DB↗

arXiv2021

PowerCut and Obfuscator: An Exploration of the Design Space for Privacy-Preserving Interventions for Voice Assistants

The pervasive use of smart speakers has raised numerous privacy concerns. While work to date provides an understanding of user perceptions of these threats, limited research focuses on how we can mitigate these concerns, either through redesigning the smart speaker or through dedicated privacy-preserving interventions. In this paper, we present the design and prototyping of two privacy-preserving interventions: `Obfuscator' targeted at disabling recording at the microphones, and `PowerCut' targeted at disabling power to the smart speaker. We present our findings from a technology probe study involving 24 households that interacted with our prototypes; the primary objective was to gain a better understanding of the design space for technological interventions that might address these concerns. Our data and findings reveal complex trade-offs among utility, privacy, and usability and stresses the importance of multi-functionality, aesthetics, ease-of-use, and form factor. We discuss the implications of our findings for the development of subsequent interventions and the future design of smart speakers.

cs.HC↗

arXiv2021

Target Set Selection parameterized by vertex cover and more

Given a simple, undirected graph $G$ with a threshold function $τ:V(G) \rightarrow \mathbb{N}$, the \textsc{Target Set Selection} (TSS) problem is about choosing a minimum cardinality set, say $S \subseteq V(G)$, such that starting a diffusion process with $S$ as its seed set will eventually result in activating all the nodes in $G$. For any non-negative integer $i$, we say a set $T\subseteq V(G)$ is a "degree-$i$ modulator" of $G$ if the degree of any vertex in the graph $G-T$ is at most $i$. Degree-$0$ modulators of a graph are precisely its vertex covers. Consider a graph $G$ on $n$ vertices and $m$ edges. We have the following results on the TSS problem: -> It was shown by Nichterlein et al. [Social Network Analysis and Mining, 2013] that it is possible to compute an optimal-sized target set in $O(2^{(2^{t}+1)t}\cdot m)$ time, where $t$ denotes the cardinality of a minimum degree-$0$ modulator of $G$. We improve this result by designing an algorithm running in time $2^{O(t\log t)}n^{O(1)}$. -> We design a $2^{2^{O(t)}}n^{O(1)}$ time algorithm to compute an optimal target set for $G$, where $t$ is the size of a minimum degree-$1$ modulator of $G$.

cs.CC↗

arXiv2021

A Social Distancing-Based Facility Location Approach for Combating COVID-19

In this paper, we introduce and study the problem of facility location along with the notion of \emph{`social distancing'}. The input to the problem is the road network of a city where the nodes are the residential zones, edges are the road segments connecting the zones along with their respective distance. We also have the information about the population at each zone, different types of facilities to be opened and in which number, and their respective demands in each zone. The goal of the problem is to locate the facilities such that the people can be served and at the same time the total social distancing is maximized. We formally call this problem as the \textsc{Social Distancing-Based Facility Location Problem}. We mathematically quantify social distancing for a given allocation of facilities and proposed an optimization model. As the problem is \textsf{NP-Hard}, we propose a simulation-based and heuristic approach for solving this problem. A detailed analysis of both methods has been done. We perform an extensive set of experiments with synthetic datasets. From the results, we observe that the proposed heuristic approach leads to a better allocation compared to the simulation-based approach.

cs.MA↗

arXiv2021

Moonshine: An Online Randomness Distiller for Zero-Involvement Authentication

Context-based authentication is a method for transparently validating another device's legitimacy to join a network based on location. Devices can pair with one another by continuously harvesting environmental noise to generate a random key with no user involvement. However, there are gaps in our understanding of the theoretical limitations of environmental noise harvesting, making it difficult for researchers to build efficient algorithms for sampling environmental noise and distilling keys from that noise. This work explores the information-theoretic capacity of context-based authentication mechanisms to generate random bit strings from environmental noise sources with known properties. Using only mild assumptions about the source process's characteristics, we demonstrate that commonly-used bit extraction algorithms extract only about 10% of the available randomness from a source noise process. We present an efficient algorithm to improve the quality of keys generated by context-based methods and evaluate it on real key extraction hardware. Moonshine is a randomness distiller which is more efficient at extracting bits from an environmental entropy source than existing methods. Our techniques nearly double the quality of keys as measured by the NIST test suite, producing keys that can be used in real-world authentication scenarios.

cs.CR↗

arXiv2021

Budgeted Influence and Earned Benefit Maximization with Tags in Social Networks

Given a social network, where each user is associated with a selection cost, the problem of \textsc{Budgeted Influence Maximization} (\emph{BIM Problem} in short) asks to choose a subset of them (known as seed users) within an allocated budget whose initial activation leads to the maximum number of influenced nodes. Existing Studies on this problem do not consider the tag-specific influence probability. However, in reality, influence probability between two users always depends upon the context (e.g., sports, politics, etc.). To address this issue, in this paper we introduce the \textsc{Tag\mbox{-}Based Budgeted Influence Maximization problem} (\emph{TBIM Problem} in short), where along with the other inputs, a tag set (each of them is also associated with a selection cost) is given, each edge of the network has the tag specific influence probability, and here the goal is to select influential users as well as influential tags within the allocated budget to maximize the influence. Considering the fact that real-world campaigns targeted in nature, we also study the \textsc{Earned Benefit Maximization} Problem in tag specific influence probability setting, which formally we call the \textsc{Tag\mbox{-}Based Earned Benefit Maximization problem} (\emph{TEBM Problem} in short). For this problem along with the inputs of the TBIM Problem, we are given a subset of the nodes as target users, and each one of them is associated with a benefit value that can be earned by influencing them. Considering the fact that different tag has different popularity across the communities of the same network, we propose three methodologies that work based on \emph{effective marginal influence gain computation}. The proposed methodologies have been analyzed for their time and space requirements.

cs.DB↗

arXiv2021

A Co\mbox{-}Operative Game Theoretic Approach for the Budgeted Influence Maximization Problem

Given a social network of users with selection cost, the \textsc{Budgeted Influence Maximization Problem} (\emph{BIM Problem} in short) asks for selecting a subset of the nodes (known as \emph{seed nodes}) within an allocated budget for initial activation to maximize the influence in the network. In this paper, we study this problem under the \emph{co\mbox{-}operative game theoretic} framework. We model this problem as a co\mbox{-}operative game where the users of the network are the players and for a group of users, the expected influence by them under the \emph{Maximum Influence Arborences} diffusion model is its utility. We call this game as \emph{BIM Game} and show this is `non-convex' and `sub-additive'. Based on the proposed game\mbox{-}theoretic model and using the solution concept called `Shapley Value', we propose an iterative algorithm for finding seed nodes. The proposed methodology is divided into mainly two broad steps: the first one is computing the approximate marginal gain in \emph{Shapley Value} for all the nodes of the network, and the second one is selecting seed nodes from the sorted list until the budget is exhausted. We also show that the proposed methodology can even be more effective when the community structure of the network is exploited. The proposed methodologies have been implemented, and an extensive set of experiments have been conducted with three publicly available social network datasets. From the experiments, we observe that the seed set selected by the proposed methodologies lead to more number of influence nodes compared to many standard and baseline methods from the literature with a reasonable computational overhead. In particular, if the community structure of the network is exploited then there is an increase upto $2 \%$ in number of influenced nodes.

cs.SI↗

arXiv2020

Face-Off: Adversarial Face Obfuscation

Advances in deep learning have made face recognition technologies pervasive. While useful to social media platforms and users, this technology carries significant privacy threats. Coupled with the abundant information they have about users, service providers can associate users with social interactions, visited places, activities, and preferences--some of which the user may not want to share. Additionally, facial recognition models used by various agencies are trained by data scraped from social media platforms. Existing approaches to mitigate these privacy risks from unwanted face recognition result in an imbalanced privacy-utility trade-off to users. In this paper, we address this trade-off by proposing Face-Off, a privacy-preserving framework that introduces strategic perturbations to the user's face to prevent it from being correctly recognized. To realize Face-Off, we overcome a set of challenges related to the black-box nature of commercial face recognition services, and the scarcity of literature for adversarial attacks on metric networks. We implement and evaluate Face-Off to find that it deceives three commercial face recognition services from Microsoft, Amazon, and Face++. Our user study with 423 participants further shows that the perturbations come at an acceptable cost for the users.

cs.CR↗

arXiv2020

Impact of Trap Filling on Carrier Diffusion in MAPbBr$_3$ Single Crystals

We present experimental evidence showing that the effective carrier diffusion length ($L_d$) and lifetime ($τ$) depend on the carrier density in MAPbBr$_3$ single crystals. Independent measurements reveal that both $L_d$ and $τ$ decrease with an increase in photo-carrier density. Scanning photocurrent microscopy is used to extract the characteristic photocurrent $I_{ph}$ decay-length parameter, $L_d$, which is a measure of effective carrier diffusion. The $L_d$ magnitudes for electrons and holes were determined to be ~ 13.3 $μ$m and ~ 13.8 $μ$m respectively. A marginal increase in uniform light bias ($\leq 5 \times 10^{15}$ photons/cm$^2$) increases the modulated photocurrent magnitude and reduces the $L_d$ parameter by a factor of two and three for electrons and holes respectively, indicating that the recombination is not monomolecular. The $L_d$ variations were correlated to the features in photoluminescence lifetime studies. Analysis of lifetime variation shows intensity-dependent monomolecular and bimolecular recombination trends with recombination constants determined to be ~ 9.3 $\times 10^6$ s$^{-1}$ and ~ 1.4 $\times 10^{-9}$ cm$^{3}$s$^{-1}$ respectively. Based on the trends of $L_d$ and lifetime, it is inferred that the sub-band-gap trap recombination influences carrier transport in the low-intensity excitation regime, while bimolecular recombination and transport dominate at high intensity.

physics.app-ph↗

arXiv2020

An Efficient Updation Approach for Enumerating Maximal $(Δ, γ)$\mbox{-}Cliques of a Temporal Network

Given a temporal network $\mathcal{G}(\mathcal{V}, \mathcal{E}, \mathcal{T})$, $(\mathcal{X},[t_a,t_b])$ (where $\mathcal{X} \subseteq \mathcal{V}(\mathcal{G})$ and $[t_a,t_b] \subseteq \mathcal{T}$) is said to be a $(Δ, γ)$\mbox{-}clique of $\mathcal{G}$, if for every pair of vertices in $\mathcal{X}$, there must exist at least $γ$ links in each $Δ$ duration within the time interval $[t_a,t_b]$. Enumerating such maximal cliques is an important problem in temporal network analysis, as it reveals contact pattern among the nodes of $\mathcal{G}$. In this paper, we study the maximal $(Δ, γ)$\mbox{-}clique enumeration problem in online setting; i.e.; the entire link set of the network is not known in advance, and the links are coming as a batch in an iterative manner. Suppose, the link set till time stamp $T_{1}$ (i.e., $\mathcal{E}^{T_{1}}$), and its corresponding $(Δ, γ)$-clique set are known. In the next batch (till time $T_{2}$), a new set of links (denoted as $\mathcal{E}^{(T_1,T_2]}$) is arrived.

cs.DS↗

arXiv2020

An Approximate Marginal Spread Computation Approach for the Budgeted Influence Maximization with Delay

In this paper, we study the Budgeted Influence Maximization with Delay Problem, for which the number of literature are limited. We propose an approximate marginal spread computation\mbox{-}based approach for solving this problem. The proposed methodology has been implemented with three benchmark social network datasets and the obtained results are compared with the existing methods from the literature. Experimental results show that the proposed approach is able to select seed nodes which leads to more number of influential nodes with reasonable computational time.

cs.SI↗

arXiv2020

One parameter family of rationally extended isospectral potentials

We start from a given one dimensional rationally extended potential associated with $X_m$ exceptional orthogonal polynomials and using the idea of supersymmetry in quantum mechanics, we obtain one continuous parameter ($λ$) family of rationally extended strictly isospectral potentials whose solutions are also associated with Xm exceptional orthogonal polynomials. We illustrate this construction by considering three well known rationally extended potentials, two with pure discrete spectrum (the extended radial oscillator and the extended Scarf-I) and one with both the discrete and the continuous spectrum (the extended generalized Poschl-Teller) and explicitly construct the corresponding one continuous parameter family of rationally extended strictly isospectral potentials. Further, in the special case of $λ= 0$ and $-1$, we obtain two new exactly solvable rationally extended potentials, namely the rationally extended Pursey and the rationally extended Abhrahm-Moses potentials respectively. We illustrate the whole procedure by discussing in detail the particular case of the $X_1$ rationally extended one parameter family of potentials including the corresponding Pursey and the Abraham Moses potentials.

quant-ph↗

arXiv2020

First Stretch then Shrink and Bulk: A Two Phase Approach for Enumeration of Maximal $(Δ, γ)$\mbox{-}Cliques of a Temporal Network

A \emph{Temporal Network} (also known as \emph{Link Stream} or \emph{Time-Varying Graph}) is often used to model a time-varying relationship among a group of agents. It is typically represented as a collection of triplets of the form $(u,v,t)$ that denotes the interaction between the agents $u$ and $v$ at time $t$. For analyzing the contact patterns of the agents forming a temporal network, recently the notion of classical \textit{clique} of a \textit{static graph} has been generalized as \textit{$Δ$\mbox{-}Clique} of a Temporal Network. In the same direction, one of our previous studies introduces the notion of \textit{$(Δ, γ)$\mbox{-}Clique}, which is basically a \textit{vertex set}, \textit{time interval} pair, in which every pair of the clique vertices are linked at least $γ$ times in every $Δ$ duration of the time interval. In this paper, we propose a different methodology for enumerating all the maximal $(Δ, γ)$\mbox{-}Cliques of a given temporal network. The proposed methodology is broadly divided into two phases. In the first phase, each temporal link is processed for constructing $(Δ, γ)$\mbox{-}Clique(s) with maximum duration. In the second phase, these initial cliques are expanded by vertex addition to form the maximal cliques. From the experimentation carried out on $5$ real\mbox{-}world temporal network datasets, we observe that the proposed methodology enumerates all the maximal $(Δ,γ)$\mbox{-}Cliques efficiently, particularly when the dataset is sparse. As a special case ($γ=1$), the proposed methodology is also able to enumerate $(Δ,1) \equiv Δ$\mbox{-}cliques with much less time compared to the existing methods.

cs.DS↗

arXiv2020

Earned Benefit Maximization in Social Networks Under Budget Constraint

Given a social network with nonuniform selection cost of the users, the problem of \textit{Budgeted Influence Maximization} (BIM in short) asks for selecting a subset of the nodes within an allocated budget for initial activation, such that due to the cascading effect, influence in the network is maximized. In this paper, we study this problem with a variation, where a set of nodes are designated as target nodes, each of them is assigned with a benefit value, that can be earned by influencing them, and our goal is to maximize the earned benefit by initially activating a set of nodes within the budget. We call this problem as the \textsc{Earned Benefit Maximization Problem}. First, we show that this problem is NP\mbox{-}Hard and the benefit function is \textit{monotone}, \textit{sub\mbox{-}modular} under the \textit{Independent Cascade Model} of diffusion. We propose an incremental greedy strategy for this problem and show, with minor modification it gives $(1-\frac{1}{\sqrt{e}})$\mbox{-}factor approximation guarantee on the earned benefit. Next, by exploiting the sub\mbox{-}modularity property of the benefit function, we improve the efficiency of the proposed greedy algorithm. Then, we propose a hop\mbox{-}based heuristic method, which works based on the computation of the `expected earned benefit' of the effective neighbors corresponding to the target nodes. Finally, we perform a series of extensive experiments with four real\mbox{-}life, publicly available social network datasets. From the experiments, we observe that the seed sets selected by the proposed algorithms can achieve more benefit compared to many existing methods. Particularly, the hop\mbox{-}based approach is found to be more efficient than the other ones for solving this problem.

cs.SI↗

arXiv2020

DySky: Dynamic Skyline Queries on Uncertain Graphs

Given a graph, and a set of query vertices (subset of the vertices), the dynamic skyline query problem returns a subset of data vertices (other than query vertices) which are not dominated by other data vertices based on certain distance measure. In this paper, we study the dynamic skyline query problem on uncertain graphs (DySky). The input to this problem is an uncertain graph, a subset of its nodes as query vertices, and the goal here is to return all the data vertices which are not dominated by others. We employ two distance measures in uncertain graphs, namely, \emph{Majority Distance}, and \emph{Expected Distance}. Our approach is broadly divided into three steps: \emph{Pruning}, \emph{Distance Computation}, and \emph{Skyline Vertex Set Generation}. We implement the proposed methodology with three publicly available datasets and observe that it can find out skyline vertex set without taking much time even for million sized graphs if expected distance is concerned. Particularly, the pruning strategy reduces the computational time significantly.

cs.DB↗