Continuous authentication with feature-level fusion of touch gestures and keystroke dynamics to solve security and usability issues

The traditional static password-based authentication schemes are insufficient to fully protect user privacy. They suffer from various usability and security issues which are usually utilized by online criminals to obtain users' passwords and control their accounts. In 2014 more than 4.9 million Gmail accounts have been compromise and published online. As a response to such incidents security researchers have started to investigates other alternative authentication methods. Currently there is a growing body of research to improve user authentication based on physiological and behavioral biometrics. In this paper we review the literature and discuss several related studies. Additionally, we propose the use of free-password authentication scheme using vein recognition at the login stage and keystroke dynamics as a continuous authentication during the user session.

Session authentication schemes establish the identity of the user only at the beginning of the session, so they are vulnerable to attacks that tamper with communications after the establishment of the authenticated session. Moreover, smartphones themselves are used as authentication means, especially in two-factor authentication schemes, which are often required by several services. Whether the smartphone is in the hands of the legitimate user constitutes a great concern, and correspondingly whether the legitimate user is the one who uses the services. In response to these concerns, Behavioral Biometrics (BB) Continuous Authentication (CA) technologies have been proposed on a large corpus of literature. This paper presents a research on the development and validation of a BBCA system (named BioPrivacy), that is based on the user’s keystroke dynamics, using a Multi-Layer Perceptron (MLP). Also, we introduce a new behavioral biometrics collection tool, and we propose a methodology for the selection of an appropriate set of behavioral biometrics. Our system achieved 97.18% Accuracy, 0.02% Equal Error Rate (EER), 97.2% True Acceptance Rate (TAR) and 0.02% False Acceptance Rate (FAR).KeywordsMachine learningBehavioral biometricsContinuous authenticationMobile devicesMulti-layer perceptron (MLP)

International audience

Continuous authentication utilizes automatic recognition of certain user\nfeatures for seamless and passive authentication without requiring user\nattention. Such features can be divided into categories of physiological\nbiometrics and behavioral biometrics. Keystroke dynamics is proposed for\nbehavioral biometrics-oriented authentication by recognizing users by means of\ntheir typing patterns. However, it has been pointed out that continuous\nauthentication using physiological biometrics and behavior biometrics incur\nprivacy risks, revealing personal characteristics and activities. In this\npaper, we consider a previously proposed keystroke dynamics-based\nauthentication scheme that has no privacy-preserving properties. In this\nregard, we propose a generic privacy-preserving version of this authentication\nscheme in which all user features are encrypted -- preventing disclosure of\nthose to the authentication server. Our scheme is generic in the sense that it\nassumes homomorphic cryptographic primitives. Authentication is conducted on\nthe basis of encrypted data due to the homomorphic cryptographic properties of\nour protocol.\n

Usability issues are vital components for online-based businesses. With Nigeria integrating electronic payment into its financial system coupled with rising internet penetration in the country, several businesses have created an online presence and are cashing in on the opportunities. This has created a form of online competition among e-commerce businesses. This study employs the user test method to test the usability issues associated with E-commerce websites in Nigeria and how this affects the success of e-commerce businesses. We find several usability issues with all e-commerce websites tested and a general need for user-focused improvement on all the websites. We also find the issues of security and trust as salient to expand the e-commerce business in Nigerians. Based on the result and analysis, recommendations on usability, data policy, security, registration and other vital issues are offered.

Digital financial platforms rely heavily on session-based authentication mechanisms that verify user identity only at discrete checkpoints, leaving systems vulnerable to post-login account takeover attacks. Continuous authentication based on behavioral biometrics has emerged as a promising solution for maintaining persistent identity assurance throughout active sessions. This work proposes a real-time behavioral biometric framework for continuous authentication using keystroke dynamics and sequential monitoring of interaction patterns. The proposed method models user-specific behavioral structure through reconstruction consistency in a low-dimensional subspace learned from enrollment data. Incoming interaction events are evaluated in real time using reconstruction error, and a sliding temporal window aggregates behavioral deviation to support stable decision making. User-specific thresholds derived from enrollment statistics enable adaptive sensitivity control, while takeover detection latency is explicitly quantified to assess operational security effectiveness. Experiments were conducted using the CMU Keystroke Dynamics benchmark under simulated session takeover scenarios. Performance was evaluated using verification metrics including false acceptance rate, false rejection rate, equal error rate, and temporal detection delay. Results demonstrate that the proposed framework achieves reliable identity discrimination while enabling rapid detection of behavioral takeover events with low latency and stable threshold behavior. The findings confirm that continuous behavioral monitoring can significantly reduce the exposure window of compromised sessions, even under unimodal biometric sensing. The proposed framework provides a computationally efficient and deployment-ready approach for strengthening real-time security in financial transaction ecosystems.

Authentication is an important security component of almost any software application. It serves as the application’s security front door by controlling access with the goal of protecting the confidentiality and integrity of the system. However, with the large variety of software applications that an end user interacts with daily, authentication is becoming a usability issue that has the potential to weaken a system’s overall security. The increasing complexity of dealing with a variety of authentication mechanisms often causes end users to develop negative security behaviours, such as writing down passwords. Moreover, some of the currently available authentication mechanisms, such as alphanumeric passwords, raise universal access issues due to both the issue of remembering a complex sequence of characters and the difficulty some individuals may have in entering that exact sequence on a keyboard or mobile device. This article proposes an authentication approach that seeks to address these usability, universal access, and security issues.

Non-Fungible Tokens (NFTs) have emerged as a way to collect digital art as well as an investment vehicle. Despite having been popularized only recently, NFT markets have witnessed several high-profile (and high-value) asset sales and a tremendous growth in trading volumes over the last year. Unfortunately, these marketplaces have not yet received much security scrutiny. Instead, most academic research has focused on attacks against decentralized finance (DeFi) protocols and automated techniques to detect smart-contract vulnerabilities. To the best of our knowledge, we are the first to study the market dynamics and security issues of the multi-billion dollar NFT ecosystem.

Since it is one of the world's most significant financial markets, the foreign exchange (Forex) market has attracted a large number of investors. Accurately anticipating the forex trend has remained a popular but difficult issue to aid Forex traders' trading decisions. It is always a question of how precise a Forex prediction can be because of the market's tremendous complexity. The fast advancement of machine learning in recent decades has allowed artificial neural networks to be effectively adapted to several areas, including the Forex market. As a result, a slew of research articles aimed at improving the accuracy of currency forecasting has been released. The Long Short-Term Memory (LSTM) neural network, which is a special kind of artificial neural network developed exclusively for time series data analysis, is frequently used. Due to its high learning capacity, the LSTM neural network is increasingly being utilized to predict advanced Forex trading based on previous data. This model, on the other hand, can be improved by stacking it. The goal of this study is to choose a dataset using the Hurst exponent, then use a two-layer stacked Long Short-Term Memory (TLS-LSTM) neural network to forecast the trend and conduct a correlation analysis. The Hurst exponent (h) was used to determine the predictability of the Australian Dollar and United States Dollar (AUD/USD) dataset. TLS-LSTM algorithm is presented to improve the accuracy of Forex trend prediction of Australian Dollar and United States Dollar (AUD/USD). A correlation study was performed between the AUD/USD, the Euro and the Australian Dollar (EUR/AUD), and the Australian Dollar and the Japanese Yen (AUD/JPY) to see how AUD/USD movement affects EUR/AUD and AUD/JPY. The model was compared with Single-Layer Long Short-Term (SL-LSTM), Multilayer Perceptron (MLP), and Complete Ensemble Empirical Mode Decomposition with Adaptive Noise–Improved Firefly Algorithm Long Short-Term Memory. Based on the evaluation metrics Mean Square Error (MSE), Root Mean Square Error, and Mean Absolute Error, the suggested TLS-LSTM, whose data selection is based on the Hurst exponent (h) value of 0.6026, outperforms SL-LSTM, MLP, and CEEMDAN-IFALSTM. The correlation analysis conducted shows both positive and negative relations between AUD/USD, EUR/AUD, and AUD/JPY which means that a change in AUD/USD will affect EUR/AUD and AUD/JPY as recorded depending on the magnitude of the correlation coefficient (r).

Today’s conventional verification system for accessing PC and handheld devices does not incorporate any post login verification. This type of system fails to address risk involved after the initial login session. With increasing personalization of computing devices, the risk extends from the erstwhile focus on malafide intervention to the issue of improper personalization of data generated out of cooperative or inadvertent intervention. Thus to ensure authenticity of the user during their entire active login period, a continuous verification is required. The current paper considers continuous unobtrusive biometric authentication technology as an approach to eliminate this problem. It also analyzes additional requirements for the unobtrusive and continuous nature of the verification system- namely unavailability of any specific modality at a given time, performance under fluctuating quality of measurable parameters and different fusion scenarios for modalities. Finally, it provides a brief framework of continuous multimodal biometric authentication using passive modalities - namely face identification, keystroke dynamics, mouse dynamics, and fingerprint which are suitable for unobtrusive continuous authentication.

MLP, CNN, LSTM and Hybrid SVM for Stock Index Forecasting Task to INDU and FTSE100

Modern decision-making in fixed income asset management benefits from intelligent systems, which involve the use of state-of-the-art machine learning models and appropriate methodologies. We conduct the first study of bond yield forecasting using long short-term memory (LSTM) networks, validating its potential and identifying its memory advantage. Specifically, we model the 10-year bond yield using univariate LSTMs with three input sequences and five forecasting horizons. We compare those with multilayer perceptrons (MLP), univariate and with the most relevant features. To demystify the notion of black box associated with LSTMs, we conduct the first internal study of the model. To this end, we calculate the LSTM signals through time, at selected locations in the memory cell, using sequence-to-sequence architectures, uni and multivariate. We then proceed to explain the states’ signals using exogenous information, for what we develop the LSTM-LagLasso methodology. The results show that the univariate LSTM model with additional memory is capable of achieving similar results as the multivariate MLP using macroeconomic and market information. Furthermore, shorter forecasting horizons require smaller input sequences and vice-versa. The most remarkable property found consistently in the LSTM signals, is the activation/deactivation of units through time, and the specialisation of units by yield range or feature. Those signals are complex but can be explained by exogenous variables. Additionally, some of the relevant features identified via LSTM-LagLasso are not commonly used in forecasting models. In conclusion, our work validates the potential of LSTMs and methodologies for bonds, providing additional tools for financial practitioners.

Taiwan, being located on a path in the west Pacific Ocean where typhoons often strike, is often affected by typhoons. The accompanying strong winds and torrential rains make typhoons particularly damaging in Taiwan. Therefore, we aimed to establish an accurate wind speed prediction model for future typhoons, allowing for better preparation to mitigate a typhoon’s toll on life and property. For more accurate wind speed predictions during a typhoon episode, we used cutting-edge machine learning techniques to construct a wind speed prediction model. To ensure model accuracy, we used, as variable input, simulated values from the Weather Research and Forecasting model of the numerical weather prediction system in addition to adopting deeper neural networks that can deepen neural network structures in the construction of estimation models. Our deeper neural networks comprise multilayer perceptron (MLP), deep recurrent neural networks (DRNNs), and stacked long short-term memory (LSTM). These three model-structure types differ by their memory capacity: MLPs are model networks with no memory capacity, whereas DRNNs and stacked LSTM are model networks with memory capacity. A model structure with memory capacity can analyze time-series data and continue memorizing and learning along the time axis. The study area is northeastern Taiwan. Results showed that MLP, DRNN, and stacked LSTM prediction error rates increased with prediction time (1–6 hours). Comparing the three models revealed that model networks with memory capacity (DRNN and stacked LSTM) were more accurate than those without memory capacity. A further comparison of model networks with memory capacity revealed that stacked LSTM yielded slightly more accurate results than did DRNN. Additionally, we determined that in the construction of the wind speed prediction model, the use of numerically simulated values reduced the error rate approximately by 30%. These results indicate that the inclusion of numerically simulated values in wind speed prediction models enhanced their prediction accuracy.

Evaluation of Deep Learning Models for Continuous Authentication Using Behavioral Biometrics

Continuous authentication has been proposed as a possible approach for passive and seamless user authentication, using sensor data comprising biometric, behavioral, and context-oriented characteristics. Since these are personal data being transmitted and are outside the control of the user, this approach causes privacy issues. Continuous authentication has security challenges concerning poor matching rates and susceptibility of replay attacks. The security issues are mainly poor matching rates and the problems of replay attacks. In this survey, we present an overview of continuous authentication and comprehensively discusses its different modes, and issues that these modes have related to security, privacy, and usability. A comparison of privacy-preserving approaches dealing with the privacy issues is provided, and lastly recommendations for secure, privacy-preserving, and user-friendly continuous authentication.