Search

Passwords have been long used as the primary authentication method for web services. Weak passwords used by the users have prompted the use of password management tools and two-factor authentication to ensure better account security. While prior studies have studied their adoption individually, none of these studies focuses particularly on the Indian setting, which is culturally and economically different from the countries in which these studies have been done in the past. To this end, we conducted a survey with 90 participants residing in India to better understand the mindset of people on using password managers and two-factor authentication (2FA). Our findings suggest that a majority of the participants have used 2FA and password managers in some form, although they are sometimes unaware of their formal names. While many participants used some form of 2FA across all their accounts, browser-integrated and device-default password managers are predominantly utilized for less sensitive platforms such as e-commerce and social media rather than for more critical accounts like banking. The primary motivation for using password managers is the convenience of auto-filling. However, some participants avoid using password managers due to a lack of trust in these tools. Notably, dedicated third-party applications show low adoption for both password manager and 2FA. Despite acknowledging the importance of secure password practices, many participants still reuse passwords across multiple accounts, prefer shorter passwords, and use commonly predictable password patterns. Overall, the study suggests that Indians are more inclined to choose default settings, underscoring the need for tailored strategies to improve user awareness and strengthen password security practices., Comment: Submitted on 2024-10-25. Accepted at BuildSec 2024, to be published in IEEE Xplore. Original version: 8 pages

General-purpose automatic speech recognition (ASR) systems do not always perform well in goal-oriented dialogue. Existing ASR correction methods rely on prior user data or named entities. We extend correction to tasks that have no prior user data and exhibit linguistic flexibility such as lexical and syntactic variations. We propose a novel context augmentation with a large language model and a ranking strategy that incorporates contextual information from the dialogue states of a goal-oriented conversational AI and its tasks. Our method ranks (1) n-best ASR hypotheses by their lexical and semantic similarity with context and (2) context by phonetic correspondence with ASR hypotheses. Evaluated in home improvement and cooking domains with real-world users, our method improves recall and F1 of correction by 34% and 16%, respectively, while maintaining precision and false positive rate. Users rated .8-1 point (out of 5) higher when our correction method worked properly, with no decrease due to false positives., Comment: Accepted to COLING 2025 Industry Track

Plant diseases significantly impact our food supply, causing problems for farmers, economies reliant on agriculture, and global food security. Accurate and timely plant disease diagnosis is crucial for effective treatment and minimizing yield losses. Despite advancements in agricultural technology, a precise and early diagnosis remains a challenge, especially in underdeveloped regions where agriculture is crucial and agricultural experts are scarce. However, adopting Deep Learning applications can assist in accurately identifying diseases without needing plant pathologists. In this study, the effectiveness of various computer vision models for detecting paddy diseases is evaluated and proposed the best deep learning-based disease detection system. Both classification and detection using the Paddy Doctor dataset, which contains over 20,000 annotated images of paddy leaves for disease diagnosis are tested and evaluated. For detection, we utilized the YOLOv8 model-based model were used for paddy disease detection and CNN models and the Vision Transformer were used for disease classification. The average mAP50 of 69% for detection tasks was achieved and the Vision Transformer classification accuracy was 99.38%. It was found that detection models are effective at identifying multiple diseases simultaneously with less computing power, whereas classification models, though computationally expensive, exhibit better performance for classifying single diseases. Additionally, a mobile application was developed to enable farmers to identify paddy diseases instantly. Experiments with the app showed encouraging results in utilizing the trained models for both disease classification and treatment guidance., Comment: 21 pages,12 figures and 2 tables

In this article, we study the effect of invisible neutrino decay of the third neutrino state for accelerator neutrino experiments at two different baselines, 1300 km with a liquid argon time projection chamber (LArTPC) detector (similar to DUNE) and 2588 km with a water Cherenkov detector (similar to P2O). For such baselines, the matter effect starts to become important. Our aim is to ascertain the sensitivity to mass hierarchy and octant of $\theta_{23}$ in these two experiments in the presence of a decaying neutrino state. We compare and contrast the results of the two experimental setups. We find that, in general, hierarchy sensitivity decreases in the presence of decay. However, if we consider decay only in the opposite hierarchy (test scenario), in the 2588 km setup, the hierarchy sensitivity with the true hierarchy as IH is larger than the no decay case. We also study the dependence of hierarchy sensitivity with true $\theta_{23}$. We find that the dominant muon background in P2O plays an important role in how the hierarchy sensitivity depends on $\theta_{23}$. The octant sensitivity for both setups increases in the presence of decay except for the LArTPC setup in case true $\theta_{23}=49^\circ$. To understand the octant sensitivity results in the two setups, we check the synergy in sensitivity between electron and muon channels as a function of test $\theta_{23}$. We also study the degeneracies in the test $\theta_{23}-\delta_{CP}$ plane and find that combined analysis of the two setups removes all the degeneracies in the test $\theta_{23}-\delta_{CP}$ plane at $5\sigma$ significance., Comment: v1: 24 pages, 12 captioned figures, 3 tables

Friction modulation allows for a range of different sensations and textures to be simulated on flat touchscreens, yet is largely unable to render fundamental tactile interactions such as path following or shape discrimination due to lack of spatial force distribution across the fingerpad. In order to expand the range of sensations rendered via friction modulation, in this paper we explore the possibility of applying spatial feedback on the fingerpad via differing friction forces on flat touchscreens. To this end, we fabricated six distinct flat surfaces with different spatial distributions of friction and observed deformation of the fingerpad skin in response to motion along these physical samples. In our study, friction changes that occur sequentially along the sliding direction introduced little transitory spatial warping such as compression or stretching to the fingerpad, suggesting limited perceptual differences in comparison to 'classic' friction modulation. Distributing friction across the direction of motion, however, showed pattern-dependent shearing of the fingertip skin, opening avenues for new sensations and illusions heretofore unachievable on flat touchscreen surfaces., Comment: 6 pages, short journal paper

The measurement of the electronic structure of metal alloys is hampered by the in-applicability of conventional quantum oscillation measurements to high-temperatures and alloy scattering. Recent advancements in the study of the electronic contribution to the entropy suggest a path to ground transport properties in alloy equilibrium thermodynamics. Fe-Ni represents an interesting test-case because it exhibits an intricate electronic structure, order-disorder transformations, and a large solid-solution region where equilibrium data can be obtained. Using cluster modeling, the electronic contribution to the entropy can be inferred from high-temperature thermodynamic data. Electronic transport property measurements can be used to independently evaluate the electronic contribution to the entropy. Reconciling these two approaches at high temperature confirms a new path to study electronic structure for metal alloys.

While witnessing the exceptional success of machine learning (ML) technologies in many applications, users are starting to notice a critical shortcoming of ML: correlation is a poor substitute for causation. The conventional way to discover causal relationships is to use randomized controlled experiments (RCT); in many situations, however, these are impractical or sometimes unethical. Causal learning from observational data offers a promising alternative. While being relatively recent, causal learning aims to go far beyond conventional machine learning, yet several major challenges remain. Unfortunately, advances are hampered due to the lack of unified benchmark datasets, algorithms, metrics, and evaluation service interfaces for causal learning. In this paper, we introduce {\em CausalBench}, a transparent, fair, and easy-to-use evaluation platform, aiming to (a) enable the advancement of research in causal learning by facilitating scientific collaboration in novel algorithms, datasets, and metrics and (b) promote scientific objectivity, reproducibility, fairness, and awareness of bias in causal learning research. CausalBench provides services for benchmarking data, algorithms, models, and metrics, impacting the needs of a broad of scientific and engineering disciplines.

EuM2As2 (M = Zn, Cd, In, Sn etc.) is an excellent material system for studying topological properties, which can be easily tuned by magnetism involved. Theoretical calculations predict gapped and flat bands in EuZn2As2 but gapless structure in EuCd2As2. In this work, low-temperature (77 K) cleaved EuZn2As2 crystals are studied using scanning tunneling microscopy/spectroscopy (STM/S) and density functional theory (DFT) calculations. Defects-induced local density of states (LDOS) modification with a triangular shape helps identify the surface terminations: Eu versus AsZn surface. While large bandgaps (~1.5 eV at 77 K) are observed on both pristine surfaces, the bandgap width is found to be very sensitive to local heterogeneity, such as defects and step edges, with the tendency of reduction. Combining experimental data with DFT simulations, we conclude that the modified bandgap in the heterogeneous area arises from Zn vacancies and/or substitution by As atoms. Our investigation offers important information for reevaluating the electron topology of the EuM2As2 family., Comment: 17 pages, 4 figures

Continuous Wave Laser Beam (LB) melting offers control over the localized heating and cooling of melt pools for welding, brazing, additive manufacturing, and solidification. Research into laser-liquid metal interactions have primarily focused on the heat and mass transport under large thermal gradients imposed by the localized melting conditions. However, little research has been conducted into varying input laser frequency and scan rate at fixed absorption to understand the effects of laser-light on surface oxide formation. This article conducts laser melting of 304L under Blue (450 nm) and Infrared (IR, 1064 nm) laser frequencies and examines their impact on laser oxide thickness, chemistry, and coloration. We find that laser frequencies induce changes in the oxide layer thickness and chemistry that cannot be explained using conventional thermal absorption shifts and fluid dynamics. We suggest that light coupling may have thermodynamic implications for the chemical potential of the liquid metal which may drive the observed phase behavior.

This paper explores recommender systems in social networks which leverage information such as item rating, intra-item similarities, and trust graph. We demonstrate that item-rating information is more influential than other information types in a collaborative filtering approach. The trust graph-based approaches were found to be more robust to network adversarial attacks due to hard-to-manipulate trust structures. Intra-item information, although sub-optimal in isolation, enhances the consistency of predictions and lower-end performance when fused with other information forms. Additionally, the Weighted Average framework is introduced, enabling the construction of recommendation systems around any user-to-user similarity metric.

Probing magnetic fields in high-redshift galactic systems is crucial to investigate galactic dynamics and evolution. Utilizing the rotation measure of the background quasars, we have developed a radial profile of the magnetic field in a typical high-$z$ galaxy. We have compiled a catalog of 59 confirmed quasar sightlines, having one intervening Mg \rom{2} absorber in the redshift range $0.372\leq z_{\text{abs}} \leq 0.8$. The presence of the foreground galaxy is ensured by comparing the photometric and spectroscopic redshifts within $3 \sigma_{z-\text{photo}}$ and visual checks. These quasar line-of-sights (LoS) pass through various impact parameters (D) up to $160$ kpc, covering the circumgalactic medium of a typical Milky-Way type galaxy. Utilizing the residual rotation measure (RRM) of these sightlines, we estimated the excess in RRM dispersion, $\sigma_{\text{ex}}^{\text{RRM}}$. We found $\sigma_{\text{ex}}^{\text{RRM}}$ decreases with increasing D. We translated $\sigma_{\text{ex}}^{\text{RRM}}$ to average LoS magnetic field strength, $\langle B_{\|}\rangle$ by considering a typical electron column density. Consequently, the decreasing trend is sustained in the magnetic field. In particular for sightlines with $\text{D} \leq 50$ kpc and $\text{D} > 50$ kpc, $\langle B_{\|}\rangle$ is found to be $2.39 \pm 0.7 \ \mu$G and $1.67 \pm 0.38 \ \mu$G, respectively. This suggests a clear indication of varying magnetic field from the disk to the circumgalactic medium. This work provides a methodology that, when applied to ongoing and future radio polarisation surveys such as LOFAR and SKA, promises to significantly enhance our understanding of magnetic field mapping in galactic systems., Comment: Published in JCAP

Content moderation faces a challenging task as social media's ability to spread hate speech contrasts with its role in promoting global connectivity. With rapidly evolving slang and hate speech, the adaptability of conventional deep learning to the fluid landscape of online dialogue remains limited. In response, causality inspired disentanglement has shown promise by segregating platform specific peculiarities from universal hate indicators. However, its dependency on available ground truth target labels for discerning these nuances faces practical hurdles with the incessant evolution of platforms and the mutable nature of hate speech. Using confidence based reweighting and contrastive regularization, this study presents HATE WATCH, a novel framework of weakly supervised causal disentanglement that circumvents the need for explicit target labeling and effectively disentangles input features into invariant representations of hate. Empirical validation across platforms two with target labels and two without positions HATE WATCH as a novel method in cross platform hate speech detection with superior performance. HATE WATCH advances scalable content moderation techniques towards developing safer online communities.

This research endeavors to address the pervasive issue of light pollution through an interdisciplinary approach, leveraging data science and machine learning techniques. By analyzing extensive datasets and research findings, we aim to develop predictive models capable of estimating the degree of sky glow observed in various locations and times. Our research seeks to inform evidence-based interventions and promote responsible outdoor lighting practices to mitigate the adverse impacts of light pollution on ecosystems, energy consumption, and human well-being.

Atomic arrangements and local sub-structures fundamentally influence emergent material functionalities. The local structures are conventionally probed using spatially resolved studies and the property correlations are usually deciphered by a researcher based on sequential explorations and auxiliary information, thus limiting the throughput efficiency. Here we demonstrate a Bayesian deep learning based framework that automatically correlates material structure with its electronic properties using scanning tunneling microscopy (STM) measurements in real-time. Its predictions are used to autonomously direct exploration toward regions of the sample that optimize a given material property. This autonomous method is deployed on the low-temperature ultra-high vacuum STM to understand the structure-property relationship in a europium-based semimetal, EuZn2As2, one of the promising candidates for studying the magnetism-driven topological properties. The framework employs a sparse sampling approach to efficiently construct the scalar-property space using a minimal number of measurements, about 1 - 10 % of the data required in standard hyperspectral imaging methods. We further demonstrate a target-property-guided active learning of structures within a multiscale framework. This is implemented across length scales in a hierarchical fashion for the autonomous discovery of structural origins for an observed material property. This framework offers the choice to select and derive a suitable scalar property from the spectroscopic data to steer exploration across the sample space. Our findings reveal correlations of the electronic properties unique to surface terminations, local defect density, and point defects.

Thanks to the large cross-section for the production of strange and charm mesons at the Large Hadron Collider, LHCb has recently performed several new explorations of rare transitions that are highly suppressed or forbidden by the Standard Model: mostly flavour changing neutral current decays, but also lepton flavour violating decays. Physics beyond the Standard Model could contribute at tree and loop levels., Comment: Presented at the 12th Workshop on the CKM Unitarity Triangle, 18-22 September 2023, Santiago de Compostela

Although social media platforms are a prominent arena for users to engage in interpersonal discussions and express opinions, the facade and anonymity offered by social media may allow users to spew hate speech and offensive content. Given the massive scale of such platforms, there arises a need to automatically identify and flag instances of hate speech. Although several hate speech detection methods exist, most of these black-box methods are not interpretable or explainable by design. To address the lack of interpretability, in this paper, we propose to use state-of-the-art Large Language Models (LLMs) to extract features in the form of rationales from the input text, to train a base hate speech classifier, thereby enabling faithful interpretability by design. Our framework effectively combines the textual understanding capabilities of LLMs and the discriminative power of state-of-the-art hate speech classifiers to make these classifiers faithfully interpretable. Our comprehensive evaluation on a variety of English language social media hate speech datasets demonstrate: (1) the goodness of the LLM-extracted rationales, and (2) the surprising retention of detector performance even after training to ensure interpretability. All code and data will be made available at https://github.com/AmritaBh/shield., Comment: Camera-ready for NAACL WOAH 2024 (Workshop on Online Abuse and Harms). First two authors contributed equally

Savonius turbines, prominent in small-scale wind turbine applications operating under low-speed conditions, encounter limitations due to opposing torque on the returning blade, impeding high efficiency. A viable solution involves mitigating this retarding torque by directing incoming airflow through a cylindrical deflector. However, such flow control is highly contingent upon the location and size of the cylindrical deflector, and its angular velocity. This study introduces a novel design optimization framework tailored for enhancing the turbine-deflector system's performance. Leveraging surrogate models for computational efficiency, six different models were assessed, with Kriging selected for subsequent analysis based on its superior performance at approximating the relation between design parameters and objective function. The training data for the surrogate model and the flow field data around the system were obtained through Unsteady Reynolds-Averaged Navier Stokes (URANS) simulations using a sliding mesh technique. An in-house code for the Quantum-based Salp Swarm Optimization (QSSO) algorithm was then employed to obtain design parameters corresponding to the peak power coefficient (Cp) for the stationary deflector-turbine system. Additionally, the QSSO algorithm was quantitatively compared with nine other competing algorithms. The optimized stationary deflector-turbine system showed an improvement of 26.94% in Cp at Tip Speed Ratio (TSR) of 0.9 compared to the baseline case. Further investigation into the effect of deflector rotational velocity ($\omega_d$) revealed significant improvements: 40.98% and 11.33% enhancement at $\omega_d$ = 3 rad/s, and 51.23% and 19.42% at $\omega_d$ = 40 rad/s, compared to configurations without a deflector and with the optimized stationary deflector, respectively at a TSR of 0.9., Comment: 42 pages, 16 figures

We have witnessed lately a rapid proliferation of advanced Large Language Models (LLMs) capable of generating high-quality text. While these LLMs have revolutionized text generation across various domains, they also pose significant risks to the information ecosystem, such as the potential for generating convincing propaganda, misinformation, and disinformation at scale. This paper offers a review of AI-generated text forensic systems, an emerging field addressing the challenges of LLM misuses. We present an overview of the existing efforts in AI-generated text forensics by introducing a detailed taxonomy, focusing on three primary pillars: detection, attribution, and characterization. These pillars enable a practical understanding of AI-generated text, from identifying AI-generated content (detection), determining the specific AI model involved (attribution), and grouping the underlying intents of the text (characterization). Furthermore, we explore available resources for AI-generated text forensics research and discuss the evolving challenges and future directions of forensic systems in an AI era.

Machine Learning (ML) has become an integral aspect of many real-world applications. As a result, the need for responsible machine learning has emerged, focusing on aligning ML models to ethical and social values, while enhancing their reliability and trustworthiness. Responsible ML involves many issues. This survey addresses four main issues: interpretability, fairness, adversarial robustness, and domain generalization. Feature selection plays a pivotal role in the responsible ML tasks. However, building upon statistical correlations between variables can lead to spurious patterns with biases and compromised performance. This survey focuses on the current study of causal feature selection: what it is and how it can reinforce the four aspects of responsible ML. By identifying features with causal impacts on outcomes and distinguishing causality from correlation, causal feature selection is posited as a unique approach to ensuring ML models to be ethically and socially responsible in high-stakes applications.

Reinforcement learning has been demonstrated to outperform even the best humans in complex domains like video games. However, running reinforcement learning experiments on the required scale for autonomous driving is extremely difficult. Building a large scale reinforcement learning system and distributing it across many GPUs is challenging. Gathering experience during training on real world vehicles is prohibitive from a safety and scalability perspective. Therefore, an efficient and realistic driving simulator is required that uses a large amount of data from real-world driving. We bring these capabilities together and conduct large-scale reinforcement learning experiments for autonomous driving. We demonstrate that our policy performance improves with increasing scale. Our best performing policy reduces the failure rate by 64% while improving the rate of driving progress by 25% compared to the policies produced by state-of-the-art machine learning for autonomous driving.