User Keystrokes Research Articles

GazeLink: A multi-language low-cost mobile eye-gesture communication system with large language models for people with amyotrophic lateral sclerosis

Amyotrophic Lateral Sclerosis (ALS) patients who have severe motor and speech impairments mostly rely on their eyes and assistive technology to communicate. However, existing high-tech products are expensive and hard to access, while low-tech products are inefficient and restrictive. To mitigate the limitations, this research proposes GazeLink, a multi-language low-cost mobile application for ALS patients to communicate efficiently with only eye movements. First, the system recognizes user eye gestures like left or up with machine learning and a template-matching algorithm. Then, it converts the eye gestures to words through a keyboard that supports English, Spanish, and Chinese. For efficiency, the system employs Large Language Models (LLMs) to generate a suitable sentence with words typed by the user and the context. Finally, the system provides text-to-speech and social media post services for both verbal and digital eye-gesture communication. Simulations conclude that sentence generation with LLMs can reduce user keystrokes by 81% while maintaining 90% of semantic similarity. Usability studies with 30 participants show that GazeLink can recognize eye gestures with 94.1% accuracy in varying lighting. After rapidly learning the user interface in under 10 attempts, first-time participants typed sentences of various lengths with their eyes at 15.1 words per minute, which is 7.2x faster than the common low-tech solution E-Tran. Experiments demonstrate GazeLinks efficiency, learnability, and accuracy in eye-gesture text entry. The system is extremely affordable (less than $0.1 a month), portable, and easily accessible online. It also supports different users, lighting, smartphones, and languages. Product testing with ALS patients and personalized LLM models will be the next step.

Integrating quantitative and qualitative analysis to evaluate digital applications in museums

Analyzing the effectiveness of a digital museum application is generally a complex process. This is due to the interaction between a digital tool capable of a limited number of narration paths and a human being whose multifaceted reactions are uniquely affected by many cultural, psychological, and emotional elements. The classical approaches in the literature can be divided into two main areas: qualitative methods, usually based on interviews at the end of the experience for assessing the cognitive effect of the digital interaction, and quantitative approaches, generally more oriented with the assessment of the subjects’ emotional reactions. Among the quantitative methods, we can mention those dealing with measurements of physiological parameters during the experience, such as ECG, EEG, temperature, etc. This paper proposes a novel approach based on integrating a qualitative process using a two-stage interview, with the quantitative analysis of the user's keystrokes during his digital interaction with the application. The results show how the quantitative component can decouple the interview answers from the interviewer's possible interferences, simultaneously delivering: i) a more transparent assessment of the user's learning experience and ii) mapping of the application's least effective sections. This provides crucial input for improving the design of digital applications for museums.

Bimodal Keystroke Dynamics-Based Authentication for Mobile Application Using Anagram

Currently, most of the smartphones recognize uses based on static biometrics, such as face and fingerprint. However, those traits were vulnerable against spoofing attack. For overcoming this problem, dynamic biometrics like the keystroke and gaze are introduced since it is more resistant against spoofing attack. This research focuses on keystroke dynamics for strengthening the user recognition system against spoofing attacks. For recognizing a user, the user keystrokes feature used in the login process is compared with keystroke features stored in the keystroke features database. For evaluating the accuracy of the proposed system, words generated based on the Indonesian anagram are used. Furthermore, for conducting the experiment, 34 participants were asked to type a set of words using the smartphone keyboard. Then, each user’s keystroke is recorded. The keystroke dynamic feature consists of latency and digraph which are extracted from the record. According to the experiment result, the error of the proposed method is decreased by 23.075% of EER with FAR and FRR are decreased by 16.381% and 10.41% respectively, compared with Kim’s method. It means that the proposed method is successful increase the biometrics performance by reducing the error rates

QIIME 2 Automation for Microbiome Analysis: An Opportunity to Reduce Barriers to Entry While Improving Analytic Speed and Reliability

The cost of next-generation sequencing has significantly reduced since its arrival in 2004, allowing metagenomic research to flourish with applications in medicine, genetics, agriculture, and environment. QIIME 2 is an open-source microbiome analysis software package that converts raw sequence data into interpretable visualizations and statistical results. Most common analyses include classifying sequences taxonomically, analyzing alpha and beta diversity, assessing phylogenetic relationship between features, and identifying features with differential abundances in various treatment groups. Third parties can contribute functionality such as longitudinal analysis and machine-learning analysis, making it a robust microbiome tool that is widely used. However, there exist several barriers in using QIIME 2. There is a steep learning curve for users with limited technical computer skill. The high volume of user inputs increases risk for user error and inefficient reproducibility among collaborators. However, these issues with efficiency, accessibility, and consistency can be addressed through automation. Using LINUX shells, we developed a robust QIIME 2 automation pipeline that automates the core functions of QIIME 2 metagenomic analysis. The process covers raw sequence data importing, demultiplexing, and denoising. It also automates data subset filtering, taxonomic classification, and alpha and beta diversity analyses. Files created with standardized nomenclature are organized into a simple folder structure comprising of data, visualizations, and metadata. QIIME 2 parameters are saved into accessible text files for transparency. The pipeline was trialed with 3 novices to both data science and metagenomic research: one PhD principal investigator, one 2nd year medical student and one 1st year medical student. They were given our scripts and tutorial, and they recorded the time and keystrokes to reach 3 endpoints: 1) construct a PCOA plot of an entire dataset, 2) perform ADONIS test on a distance matrix subset, and 3) compare alpha diversity using the Shannon metric on a metadata subset feature table. 2 out of 3 trials were successful in meeting these endpoints within 2 weeks, requiring 7 (PhD) and 9 hours (2nd year medical student) to do so. Using the fewest possible keystrokes, this analysis would have taken 11,594 characters over roughly 40 hours if done manually. Using our QIIME 2 automation, these endpoints were completed with 228 user keystrokes. The limitations of QIIME 2 can be diminished by automation and clarification of steps at decision points within a metagenomic analysis. The goal of this project was to reduce risk of error and time spent, limiting the obstacles facing a new user using QIIME 2. With growing popularity in metagenomic research, our automation tool can help computer novices navigate through the technical barriers of metagenomic analysis.

High Security User Authentication Enabled by Piezoelectric Keystroke Dynamics and Machine Learning

With the rapid development of electric and information technologies, security issues are becoming a pressing concern. This has given rise to a plethora of authentication techniques, leading among them being the fixed-password method due to its simple mechanism and independence from any specific hardware in the modern smartphone. While fixed-password methods enjoy benefits such as ease of use, potential security issues associated with password leakage cannot be ignored. This article, investigates an alternative strategy based on user keystrokes. Here, the user touch times and force features are extracted from a piezoelectric force touch panel which is an integral part of the hardware. Three broadly adopted machine learning classifiers are used for the collected data set, finally achieving an Equal Error Rate (EER) of 0.720%.

Machine Learning Methods in Authentication Problems Using Password Keystroke Dynamics

We examine the problem of user authentication from keystroke dynamics. A new static authentication method that collects information about user keystrokes is described. Its applicability to the authentication problem is substantiated by experiments and the optimal conditions for the implementation of the method are chosen.

Enhancing Keylogger Detection Performance of the Dendritic Cell Algorithm by an Enticement Strategy

Evasive software keyloggers hide their malicious behaviors to defeat run-time detection. In this paper, based on the analysis of the evasion mechanisms used by common software keyloggers, we established a framework for their detection. Using an enticement strategy, the framework we built could induce keyloggers exhibited more obvious malicious activities by mimicking user keystrokes. These 'amplified' activities are then correlated by the dendritic cell algorithm (an immune-inspired algorithm) to final determine the existence of a keylogger in a host. Preliminary experimental results showed that the framework could improve the performance of keylogger detection and hard to evade. Index Terms—keylogger, keystroke simulation, dendritic cell algorithm (DCA), correlation

Implementation of the AutoComplete Feature of the Textbox Based on Ajax and Web Service

Textbox input is widely used on the web for user input. In order to provide convenience for the web users and increase the input efficiency, one can use AutoComplete feature with the textbox. AutoComplete feature predicts possible word matches for entries that begin with the prefix typed into the textbox without the user actually typing the whole word completely. This paper describes the design and implementation of an ASP.NET application based on AutoComplete Extender control from AJAX Control Toolkit. The textbox control is bound with the AutoComplete Extender and then web service is deployed to connect the database server that stores the possible words. User keystrokes are monitored and textbox input is completed based on first typed letters that match with data stored in database. An AutoComplete test program is developed and tested.

On the Best Sensor for Keystrokes Inference Attack on Android

Abstract One of the most recently exposed security threats on smartphone platforms is the potential use of motion sensors to infer user keystrokes. Exploited as side channels, few researchers have demonstrated the ability of built-in accelerometers and gyroscopes in particular, to reveal information related to user input, though the practicality of such an attack remains an open question. This paper takes further steps along the path of exploring the aspects of the new threat, addressing the question of which available sensors can perform best in the context of the inference attack. We design and implement a benchmark experiment, against which the performances of several commodity smartphone-sensors are compared, in terms of inference accuracy. All available Android motion sensors are considered through different settings provided by the OS, and we add the option of fusing several sensors input into a single dataset, to examine the amount/lack of improvement in the attack accuracy. Our results indicate an outstanding performance of the gyroscope sensor, and the potential improvement obtained out of sensors data fusion. On the other hand, it seems that sensors with magnetometer component or the accelerometer alone have less benefit in the adverted attack.

Automatic Removal of Typed Keystrokes From Speech Signals

Computers are increasingly being used to capture audio in various applications such as video conferencing and meeting recording. In many of these applications, user may be simultaneously typing on the keyboard, e.g., to take notes or search for information. As a result, the captured speech signals are significantly corrupted by sounds generated by the user's keystrokes. In this paper we propose an algorithm to automatically detect and remove keystrokes from speech signals. The proposed method does not require any user training or enrollment and is computationally efficient. The keystroke removal algorithm generates significantly enhanced speech as measured by both user listening tests and speech recognition experiments

Evaluation of an Automated Airport Ground Transportation Information System

Described are the results of an evaluation study of a demonstration program sponsored by the Division of Aeronautics of the California Department of Transportation, in which computer-based ground transportation information kiosks were installed in four California airports, in conjunction with a similar program undertaken by the Los Angeles Department of Airports. The study included an analysis of kiosk-use patterns derived from the logfiles of user keystrokes and various system actions, which are automatically maintained by the kiosks, and surveys of both kiosk users and airport users in general to determine traveler awareness of the new information kiosks, identify passenger ground transportation information needs, evaluate user satisfaction with the information provided, and determine general travel characteristics of kiosk users. Finally, the costs involved in routine operation and in maintenance of the kiosks are examined and a number of recommendations are made on the future use of the kiosks and on the need to preserve the information generated by the demonstration program.

INTMOV: a program for the interactive analysis of spatial data

Abstract A set of programs have been developed to compute moving-windows statistics from spatial data, which are fully interactive, allowing the user to compute and display two plan views of the window statistics, the parent population histogram and the two histograms of the windows statistics all on one graphical display. Stationarity of a deposit can be checked instantly at differing window sizes with all the sample data or with partitioned subsets. Data can be read into the routine from a current exploration campaign or from working site-survey data, allowing comparisons between ‘projected’ and ‘ground’ truth. Screen and DXF file output of the local neighbourhood window statistics are plotted in a colour which corresponds to standard confidence bands of the parent or global distribution. Rapid identification of local anomalies is possible. Window statistics for all the neighbourhoods can be written to file at any subset number of the original sample number. Plots of the window means and standard deviations versus numbers of boreholes allow the exploration geologist to check stationarity in a retrospective way. The main aim in the development of these routines was to reduce the amount of user keystrokes needed to compute and display the maximum amount of neighbourhood information. The technique is demonstrated with examples from UK Coal Measures.

Keyboard Method for Composing Chinese Characters

A keyboard method is described that allows a user to form Chinese characters. The user's keystrokes activate internal logic, which performs the necessary scaling and positioning of the character components. An initial design using a 40-key keyboard is described, together with approximations of the character shapes produced by given keying sequences. Rough estimates of speed range from eight to 33 characters per minute.

Grouping of Printed Digits for Manual Entry

Recent studies on the spatial grouping of the digits are reviewed with emphasis on six new experiments by the author. For several tasks including keyboard entry and telephone dialing, grouping by 3's and 4's is consistently best in speed. Differences in error rate are generally statistically insignificant. If successive numbers have different groupings (user cannot predict grouping), differences between groupings may not be found. Users often state preferences for larger groups than those producing best performance. The temporal grouping of the user's keystrokes reflects the spatial grouping on the source document; however, the number of digits keyed after each look or referral to a number is not necessarily an indication of optimum grouping. All data point to grouping by 3's and 4's as best.