Web Sessions Research Articles

Integrating session recording and eye-tracking: development and evaluation of a Chrome extension for user behavior analysis

The subject of this article is the development, implementation, and evaluation of a Chrome extension designed to record and replay web sessions using integrated eye-tracking data. The goal is to develop a tool that enhances user interaction analysis by combining session recordings with eye-tracking capabilities. The tasks to be solved are as follows: create a functional Chrome extension that utilizes rrweb for session recording and WebGazer.js for eye tracking; implement features, such as session recording, replay with eye-tracking data overlays, and session management; and provide options for exporting and importing recorded sessions. The methods used are: architectural modeling using UML diagrams to design the system architecture, software engineering techniques for developing the extension, integration testing to ensure the smooth operation of combined features, and data preprocessing techniques to prevent redundancy and reduce noise in eye-tracking data. Additionally, a structured user study with detailed questionnaires combining both Likert-scale questions and open-ended responses and feedback analysis were conducted to evaluate usability and gather feedback. The following results were obtained: the extension was successfully developed and evaluated with 25 participants aged 18–35 years in a controlled environment. High usability ratings were obtained, with an average score of 4.5 out of 5 for the session recording, replay, and session management features. However, the eye-tracking feature received a lower rating of 3.8 out of 5 due to occasional inaccuracies in the eye-tracking data. The qualitative feedback indicated the usefulness of the eye-tracking feature and highlighted the need for improved data accuracy. Conclusions. The scientific novelty of this study lies in the integration of session recording and eye tracking within a Chrome extension, which represents a novel and comprehensive tool for user interaction analysis on the web. The tool's ability to capture both behavioral data and visual attention without requiring website code modifications is particularly valuable for researchers, marketing specialists, UI/UX designers, and product developers. The usability study and feedback analysis provided a clear direction for future improvements, including enhancing the eye-tracking accuracy and integrating advanced analytics and customizable reporting options. Future work will also explore the integration of machine learning algorithms to automatically analyze recorded data to provide deeper insights and actionable recommendations.

Exploiting web content semantic features to detect web robots from weblogs

Nowadays, web robots are predominantly used for auto-accessing web content, sharing almost one-third of the total web traffic and often posing threats to various web applications’ security, privacy, and performance. Detecting these robots is essential, and both online and offline methods are employed. One popular offline method is the use of weblog feature-based automated learning. However, this method alone cannot accurately identify web robots that continuously evolve and camouflage. Web content features combined with weblog features are used to detect such robots based on the assumption that human users exhibit specific interests while robots randomly navigate web pages. State-of-the-art web content-based feature methods lack the ability to generate coherent topics, which can confound the performance of classification models. Therefore, we propose a new content semantic feature extraction method that uses the LDA2Vec topic model, combining the strengths of LDA and the Word2Vec model to produce more semantically coherent topics by exploiting website content for a web session. To effectively detect web robots, web resource content semantic features are combined with log-based features in the proposed web robot detection approach. The proposed approach is evaluated in an e-commerce website access logs and content data. The F-score, balanced accuracy, G-mean, and Jaccard similarity are used for performance measures, and the coherence score metric is used to determine the number of topics for a session. Experimental results demonstrate that a combination of weblogs and content semantic features is effective in web robot detection.

Role of self-directed learning in health professions education: Essential role of medical graduates of India

To cope with the challenges in the present day’s healthcare, developing self-directed learning (SDL) skills for medical graduates is of immense importance and facilitators of learning have an important role in this aspect and sensitization of faculty is necessary. One-month Mentor-learner web sessions on SDL were arranged and thirty-one learners and five faculty members participated in the month-long sessions. The sessions were divided into five parts with different objectives. Three learners and five faculty members conducted the sessions and were allotted specific roles to smoothly conduct the process. Thread 1 made learners visit their experiences and use the views of peers in understanding SDL. Thread 2 was based on principles and types of SDL and methods of guiding students for SDL sessions. Thread 3 was a group activity to design SDL sessions and review other groups’ SDL models. Thread 4 focused on the motivation of students and the factors affecting them. Thread 5 was on reflection and incorporated a self-assessment link as well. Participants were satisfied with the way SDL dealt with details. The design of the module made the learning diverse and interesting. Through the learning, participants felt more confident about designing and implementing an SDL session when needed. Through the group activity, they gained respect for the thinking and abilities of co-learners.

Training to improve the vaccination offer: the experience of the AcToVax4NAM project

Abstract Background Developing training activities material with the aim of strengthening vaccination literacy and cultural aware approach among “Professional-FOR-health”, all health and social care workers and other professionals working in the field of the immunization of Newley Arrived Migrants (NAMs), is part of the activities of the project “Access To Vaccination for Newly Arrived Migrants”-AcToVax4NAM (Grant n.101018349, 3rdEU Health Programme). Objectives The Italian project Team, Sapienza University of Rome and Italian National Health Institute, designed the framework for training to strengthen the healthcare system Vaccination Literacy and responsiveness among Professional-FOR-health, in order to increase the capacity building skills and cultural sensitivity of the healthcare system in promoting active immunization of NAMs. The training materials were shared to all Consortium Countries (Cyprus, Germany, Greece, Italy, Malta, Poland, Romania, Spain) to be implemented in their context. Results The targets of the training activities are health professionals responsible for vaccination, professionals from other services that meet NAMs or promote vaccination (GPs/FPs), professionals active in the target area (NGOs). We designed a blended training, face to face sessions, web sessions should alternate. The topics are “Vaccine Literacy”, “Cultural awareness/migrant sensitivity”, “Health promotion tools with community-based approach” and “Entitlement to vaccination for NAMs/NAMs profile”. 4 evaluation tools were constructed: a pre- and post-test learning, an exploratory survey on the level of vaccine literacy of the organisations involved, a satisfaction questionnaire. Conclusions Each country will construct, based on the resources made available, a course that is as adherent as possible to the context of its own country and target group. This type of training can be the groundwork for the changes needed to make the NAMs vaccination process more guaranteed and inclusive. Key messages • It is important to stimulate a training process that enables health professionals to reconnect their daily experiences with a broader context of thinking, for a real empowerment process. • As learners’ cannot be seen as mere recipients of notions and information, their needs, motivations and experiences should be put right at the centre of the training process.

Implementation of Acupressure as Adjunct Therapy to Reduce Postoperative Nausea and Vomiting

<h3>Background Information</h3> Postoperative nausea and vomiting (PONV) is a common surgical issue that can result in complications such as, electrolyte imbalance, dehydration, pulmonary aspiration, and wound dehiscence. Additionally, PONV can delay hospital discharge, subsequently affecting hospital costs. Substantial research suggests that utilizing a multimodal approach combining prophylactic pharmacologic antiemetics with non-pharmacological therapies can reduce PONV, therefore decreasing the need to utilize multiple doses of rescue antiemetic therapy. Studies have also demonstrated that acupressure can shorten the time from extubation to "ready for discharge" from PACU. Acupressure can produce additive effects with minimal side effects and without potential for adverse drug reactions. <h3>Objectives of Project</h3> The objectives of the project include: (1) Provide education to nursing staff regarding the benefits of acupressure therapy for patients; (2) Educate nurses on the application of acupressure (3) Reduce patient's nausea by adding acupressure into treatment regimen <h3>Process of Implementation</h3> Six perioperative nurses designated as super-users, were instructed by a licensed acupressure provider on the utilization of acupressure for postoperative nausea and vomiting. Instruction was provided on the application of stimulation to specific points or defined pathways on the body (acupressure) that can relieve nausea. Educational sessions with the licensed acupressure provider were comprised of Zoom interactive web sessions, live in-person training, and evaluation of technique at patient bedside. Once deemed competent, the super-users provided acupressure for ambulatory PACU patients who were identified as being at high risk for PONV or for patients who were reported to be experiencing PONV. Super-users assessed severity of nausea before and after the implementation of acupressure to determine effectiveness by using the Edmonton scale. <h3>Statement of Successful Practice</h3> 45 out of 48 patients who received acupressure stated that "they felt less nauseous after acupressure". The evaluation of their Edmonton scores indicated that acupressure, along with nausea medications, reduced PONV. <h3>Implications for Advancing the Practice of Perianesthesia Nursing</h3> Acupressure is a non-invasive nursing intervention that provides a non-pharmacologic alternative in the treatment of PONV. Application of acupressure could be incorporated into new-hire orientation for perianesthesia nurses, in education for the treatment of PONV. Utilization of acupressure could decrease the use of pharmacologic therapy, reduce the postoperative complications, and ultimately improve patient outcomes.

Exploring the role of dramatics in medical education through online listserv mediated mentoring and learning web sessions

BACKGROUND:Online learning environments are becoming more frequent in teaching and learning than ever before. Asynchronous learning provides “high degree of interactivity” between participants and helps them to reflect upon their ideas/thoughts, before sharing them with others leading to more insightful responses and in-depth learning. Moderating an online discussion is an art and challenging task, which needs meticulous planning and effective execution approaches. One of the major challenges in online discussion is to ensure active participation and interest among participants, and therefore, it reiterates the need for inclusion of interactive strategies to enable it meaningful and effective. The present article is to share our experiences in moderating session on “Dramatics in Medical Education” (DIME) as a part of ML web discussion of Foundation for Advancement in International Medical Education and Research (FAIMER) at one of the regional FAIMER institutes in India.MATERIALS AND METHODS:Thirty-two FAIMER fellows participated in online discussion on DIME, supplemented with various engagement triggers including, sharing of real-life situation case scenarios, motivational quotes, competitions and rewards, script writing, script editing exercises by using Forum Theatre method, discussion on uploaded academic videos on “YouTube,” use of mnemonics, etc.RESULTS:Online discussion yielded improved knowledge and positive attitudes of participants toward DIME. Task-based exercises within authentic or realistic situations, teamwork, and interactions facilitated active learning of participants.CONCLUSIONS:Appropriate planning, well coordination, and interactive strategies can foster learning and reflective abilities as well as develop favorable perceptions of learners toward virtual mode of teaching-learning.

Measuring Web Session Security at Scale

Session management is a particularly delicate component of web applications, which might suffer from a range of severe security issues, including impersonation attacks. Unfortunately, the scope and significance of prior work on web session security in the wild are limited by the complexity of the attack surface and the challenges of automating the login process on existing websites. In the present article, we fill this gap by proposing the first comprehensive, large-scale web session security measurement based on post-login data. Our analysis is comprehensive in that it deals with all key aspects of web sessions, i.e., the login process, the logout process and the authentication cookie handling. Our automated approach analysed an extensive set of session management practices of over 6,000 sites where login was successful and authentication cookies could be automatically detected, uncovering a widespread adoption of insecure practices in the wild.

User Behavior Audit System of Real-time Web Log by Spark

With the rapid development of information technology, single sign-on authentication systems have been adopted by organizations to protect their information management systems and other service platforms. With the single sign-on method, users just log in the authentication systems once to access all the information in systems, which are more convenience, but accompanied by user permissions vulnerabilities and information security risks. Users’ browsing behaviors reflect the users’ usage habits, purposes, and work contents. Risks of information leakage can be reduced by auditing the users’ browsing behaviors. After analyzing the web logs of several service systems, we designed a user behavior audit system based on the Spark computing engine, which takes the real-time web log file as the data set. A possible preprocessing method for real-time web log data stream is proposed to extract users’ web session data. we used the Spark computing engine and the corresponding analysis to mine users’ behavior patterns. The online real-time detection and offline detection of user behavior are realized, and users are divided into three levels: normal, abnormal, and risky. The system administrators or auditors can view the calculation results of each stage online, manage the parameter configuration of the algorithm online, and adjust the user’s access right in real-time.

On the platform but will they buy? Predicting customers' purchase behavior using deep learning

A thorough understanding of online customer's purchase behavior will directly boost e-commerce business performance. Existing studies have overtly focused on purchase intention and used sales rank as a natural proxy, which however has limited business application. Additionally, intention to purchase does not necessarily convert to actual retail purchases. We aim to further our understanding of online customer's purchase behavior for an e-commerce platform by predicting the same using deep learning techniques, on a large multidimensional data sample of more than 50,000 unique web sessions. This study used two distinct sets of variables, i.e., platform engagement and customer characteristics, as key predictors of online purchases by retail customers. We further compared the predictive capability of our deep learning method with other widely used machine learning techniques for prediction, including Decision Tree, Random Forest, Support Vector Machines, and Artificial Neural Networks. We found that the deep learning technique outperformed the machine learning techniques when applied to the same dataset. These analyses will help platform designers plan for more platform engagements while simultaneously expanding the academic understanding of purchase prediction for online e-commerce platforms.

Session stitching using sequence fingerprinting for web page visits

The nature of people's web navigation has significantly changed in recent years. The advent of smartphones and other handheld devices has given rise to web users consulting websites with more than one device, or using a shared device. As a result, large volumes of seemingly disjoint data are available, which when analysed together can support decision-making. The task of identifying web sessions by linking such data back to a specific person, however, is hard. The idea of session stitching aims to overcome this by using machine learning inference to identify similar or identical users. Many such efforts use various demographic data or device-based features to train matching algorithms. However, often these variables are not available for every dataset or are recorded differently, making a streamlined setup difficult. Besides, they often result in vast feature spaces which are hard to use for actionable interpretation.In this paper, we present an alternative approach based on the fingerprinting of web pages visited by users in a single session. By learning behavioural patterns from these sequences of page visits, we obtain features that can be used for matching without requiring sensitive user-agent data such as IP, geo location, or device details as is common with other approaches. Using these sequential fingerprints does not rely on pre-defined features, but only requires the recording of web page visits, making our approach actionable. The approach is empirically tested on real-life web logs and compared with matching using regular user-agent features and state-of-the-art embedding techniques. Results in an ecommerce context show sequential features can still obtain strong performance with fewer features, facilitating decision-making on session stitching and inform subsequent related activities such as marketing or customer analysis.

An Exploit Kits Detection Approach Based on HTTP Message Graph

The exploit kits (EKs) are used by attackers to distribute malware automatically and silently. Existing approaches to EKs detection usually need to perform dynamic analysis on the content contained in the network traffic, which requires dumping all the network traffic and thus causes high detection overhead. Although some approaches detect EKs based on static analysis, they usually fail to restore the complete attack path because of the obstruction set by the attackers. In this paper, we propose an approach that can detect EKs based on only information extracted by static analysis. Our method builds a graph for web sessions and extracts features from the graph to perform EKs detection. The built graph catches important structural characteristics of the interaction during EK attacks that were not revealed in existing methods, with which EKs can be detected with high accuracy. The experiments show that our method works well in both the ground-truth datasets and the latest practical cases. Our method can also identify the malicious websites concealed in EKs, which can further improve the efficiency of analysis.

A parallel approach for web session identification to make recommendations efficient

A parallel approach for web session identification to make recommendations efficient

Identifying Actual Users in a Web Surfing Session using Tracing and Tracking

The World Wide Web (www) has become an interactive medium for millions of Internet users across the world, daily visiting occur for various purposes. However, their behavior could be monitored by various service providers or company’s website they visited for several purposes. Thus, we can see that if certain data pre-processing techniques are practiced correctly, results will be of high quality and in turn provide substantial evidence of the user and its web session. This paper provides a way of knowing user behavior on the web. It explains the use of cookies: Cookies allow you to do many things such as log on to Internet Banking, have us 'remember' your log-in preferences, when logging on, a pop-up window appears stating the service is not available and to try later. A misspelled domain name in the address line. Lotteries that charge a fee to collect your winnings, apply for products and services online, and use financial tools like our ‘find and compare’ mortgage calculators., servers, and proxy logs with some scripting to be able not only to get a user’s IP address but also get personal information like age, identity, and any other credential as long as the user puts them down while using the web.

Learning semantic information from Internet Domain Names using word embeddings

Word embeddings is a well-known set of techniques widely used in Natural Language Processing (NLP). These techniques are able to learn words’ semantic based on the distributional hypothesis which states that words that are used and occur in the same contexts tend to purport similar meanings.This paper explores the usage of word embeddings in a new scenario to create a Vector Space Model (VSM) for Internet Domain Names (DNS). Our goal is to find semantically similar domains only using information of DNS queries without any knowledge about the content of those domains.The results presented here have practical applications in many engineering activities including websites recommendations, identification of fraudulent or risky sites, parental-control systems and anomaly detection in network traffic analysis (among others).We use the distributional hypothesis to learn the semantic of domain names from users’ web navigation patterns, validating empirically that domain names that occur in the same web sessions tend to have similar semantic.We also test different word embeddings techniques: word2vec, app2vec (considering time intervals between DNS queries), and fastText (which includes sub-word information).Due to the characteristics of domain names, we found fastText as the best option for building a VSM for DNS, being 10.5% superior than word2vec with Skip-Gram which was the next best technique considering the Mean Average Precision at k (MAP@k) metric, which compares the most similar domains in our VSM with the most similar domains provided by a third party source, namely, similar sites service offered by Alexa Internet, Inc.

How Digital Word-of-Mouth Affects Consumer Decision Making: Evidence from Doctor Appointment Booking

We use detailed clickstream data on online word-of-mouth (WOM) to uncover mechanisms underlying its influence on consumer decision making. A feature launch on a major doctor appointment booking platform allows us to examine the effects of online WOM on three dimensions of a consumer’s choice process: the consideration set size, the time taken to consider alternatives (web session duration), and the geographic dispersion of the choices considered. Results indicate that the effects of WOM on decision-making processes are not monotonic but rather are contingent on the abundance of WOM (number of rated doctors) in a market. When the abundance of WOM is high, the introduction of WOM makes patients consider fewer doctors, browse for a shorter duration, and focus on doctors that are geographically more proximate. In contrast, when the abundance of WOM is low, the introduction of WOM makes patients consider more doctors, browse for longer duration, and consider doctors that are geographically more dispersed. We also find that WOM can lead to a cannibalization effect: when ratings are published, the highly rated doctors reap the benefits (in the form of increased demand) at the expense of unrated doctors. Our study contributes to the extant literature on online WOM by providing new insights into how WOM influences consumer decision making and by examining this question at a more granular level than prior work.This paper was accepted by Anandhi Bharadwaj, information systems.

Patients' User Experience of a Blended Face-to-Face and Web-Based Smoking Cessation Treatment: Qualitative Study.

BackgroundBlended web-based and face-to-face (F2F) treatment is a promising electronic health service because the strengths of one mode of delivery should compensate for the weaknesses of the other.ObjectiveThe aim of this study was to explore this compensation by examining patients’ user experience (UX) in a blended smoking cessation treatment (BSCT) in routine care.MethodsData on patients’ UX were collected through in-depth interviews (n=10) at an outpatient smoking cessation clinic in the Netherlands. A content analysis of the semantic domains was used to analyze patients’ UX. To describe the UX, the Hassenzahl UX model was applied, examining 4 of the 5 key elements of UX from a user’s perspective: (1) patients’ standards and expectations, (2) apparent character (pragmatic and hedonic attributes), (3) usage situation, and (4) consequences (appeal, emotions, and behavior).ResultsBSCT appeared to be a mostly positively experienced service. Patients had a positive-pragmatic standard and neutral-open expectation toward BSCT at the treatment start. The pragmatic attributes of the F2F sessions were mostly perceived as positive, whereas the pragmatic attributes of the web sessions were perceived as both positive and negative. For the hedonic attributes, there seemed to be a difference between the F2F and web sessions. Specifically, the hedonic attributes of the web sessions were experienced as mostly negative, whereas those of the F2F sessions were experienced as mostly positive. For the usage situation, the physical and social contexts were experienced positively, whereas the task and technical contexts were experienced negatively. Nevertheless, the consequential appeal of BSCT was positive. However, the consequential emotions and behavior varied, ultimately resulting in diverse combinations of consequential appeal, emotions, and behavior (positive, negative, and mixed).ConclusionsThis study provided insights into the UX of a blended treatment, and the results support the expectation that in a blended treatment, the strengths of one mode of delivery may compensate for the weaknesses of the other. However, in this certain setting, this is mainly achieved in only one way: F2F sessions compensated for the weaknesses of the web sessions. As a practical conclusion, this may mean that the web sessions, supported by the strengths of the F2F sessions, offer an interesting approach for further improving the blended treatment. Our theoretical findings reflect the relevance of the aspects of hedonism, such as fun, joy, or happiness in the UX, which were not mentioned in relation to the web sessions and were only scarcely mentioned in relation to the F2F sessions. Future research should further investigate the role of hedonistic aspects in a blended treatment and whether increased enjoyment of a blended treatment could increase treatment adherence and, ultimately, effectiveness.

A Parallel Approach for Web Session Identification to make Recommendation Efficient

A sequence of web pages visited by the clients over a particular timeframe is called a session. Web log mining is done to analyse the behaviour of the users, using the web access patterns. Sessions are identified as the significant part of the construction of the recommendation model. The novel part of the work makes use of backward moves made by the user, considering both the referrer url and the requested url extracted from the extended web log for session identification which is not taken into consideration in the existing heuristic-based approach. Two noteworthy issues in session identification are: 1) framing excessively numerous smaller length sessions; 2) taking longer time for identifying the sessions. In the proposed work, the length of the sessions are maximised using split and merge technique and the time taken for session identification is reduced using thread parallelisation. For efficient storage and retrieval of information the hash map data structure is used. The proposed work shows significant improvement in performance in terms of execution time, standard error, correlation coefficient and the objective value.

Analysis of Http Cookie Hijacking in the Wild

Because cookies act as the sole evidence of user identification, web sessions are especially vulnerable to attacks through session hijacking, where the server operated by a specific user sends users ' identity requests. If n &gt; 1 cookies are used to execute a session, n sub-sessions that actually run on the same website where the individual cookies are used to access part of the session's state details. Our cookie hijacking analysis shows a range of significant defects; attackers may reach Google's home address and work address and websites that are accessed by Bing or Baidu, show the entire browsing history of the user, and Yahoo enables attackers to delete the list of contacts and upload emails from the account of the consumer. For fact, e-commerce providers such as Amazon and Ebay have a limited, complete customer order background, so almost all platforms have a user name so e-mail address on their page. Ad networks like Doubleclick will also expose pages accessed by the customer. In this article, we propose to improve the latest state-of - the-art HTTP(S) session control by utilizing user fingerprint.A vast range of functionalities of the new client tracking makes session identification on the server observable and dramatically increases the threshold for attackers. Furthermore, this paper describes HTML5 and CSS capabilities for client fingerprinting and the recognition or authentication of a device by using the UserAgent list.

COVID‐19 leads to jump in people seeking MH help in NYC

It's a stressful time to be a New Yorker. People are stuck at home, worrying about their health, their family and their paycheck, Spectrum News New York 1 reported April 13. “We are seeing a lot of calls related to anxiety and stress that people are experiencing today,” said Matt Kudish, executive director of the National Alliance on Mental Illness of New York City. The organization says the number of callers seeking help has jumped 60% over the last few weeks, and that conversations now last about 15 minutes longer than average. The city says it's also seeing an increase in demand. Web sessions to its NYC Well service jumped 185% last month to more than 81,000. “We're hearing from people who are worried about their own physical health, they're worried about their mental health being forced to stay socially distant from other people,” Kudish said. Health care providers predict that even when the COVID‐19 crisis ends, the mental health crisis will continue. They say people will continue to be anxious — this time about leaving home and being convinced it's safe to be outside.

Covert Device Association Among Colluding Apps via Edge Processor Workload

While thriving application (app) distribution systems involving incentivized third-party app vendors are desirable for the emerging edge computing paradigm, they also bring security challenges as faced by the current mobile app distribution systems. This article studies a threat called covert device association , in which the vendors of two apps collude to figure out which of their app installations run on the same edge device. The threat can widely spread when the two apps are popular. It is also a stepping stone for: 1) the de-anonymization attacks against the users anonymous to one of the two vendors and 2) privilege escalation in which the two colluding vendors have united privileges. We show that the threat can be implemented via a reliable and ubiquitous covert channel based on the edge device’s processor workload without requiring any privileged permissions. We present the implementation details for three attack scenarios of: 1) two Android apps; 2) an Android app and a Web session running in the mobile Tor browser; and 3) two Android Things apps. Evaluation on two smartphones and an embedded edge device shows that the covert channel gives at least 0.25 b/s data rate with zero empirical bit error rate and the covert device association can be completed within 3.2 min.