The Role of Self-Improving Tutoring Systems in Fostering Pre-Service Teacher Self-Regulated Learning

Computer-based learning environments serve as a valuable asset to help strengthen teacher preparation and preservice teacher self-regulated learning. One of the most important advantages is the opportunity to collect ambient data unobtrusively as observable indicators of cognitive, affective, metacognitive, and motivational processes that mediate learning and performance. Ambient data refers to teacher interactions with the user interface that include but are not limited to timestamped clickstream data, keystroke and navigation events, as well as document views. We review the claim that computers designed as metacognitive tools can leverage the data to serve not only teachers in attaining the aims of instruction, but also researchers in gaining insights into teacher professional development. In our presentation of this claim, we review the current state of research and development of a network-based tutoring system called nBrowser, designed to support teacher instructional planning and technology integration. Network-based tutors are self-improving systems that continually adjust instructional decision-making based on the collective behaviors of communities of learners. A large part of the artificial intelligence resides in semantic web mining, natural language processing, and network algorithms. We discuss the implications of our findings to advance research into preservice teacher self-regulated learning.

The politics of piracy: political ideology and the usage of pirated online media

There is a lack of clarity in information systems research on which factors lead people to use or not use technologies of varying degrees of perceived legality. To address this gap, we use arguments from the information systems and political ideology literatures to theorize on the influence of individuals’ political ideologies on online media piracy. Specifically, we hypothesize that individuals with a more conservative ideology, and thus lower openness to experience and higher conscientiousness, generally engage in less online media piracy. We further hypothesize that this effect is stronger for online piracy technology that is legally ambiguous. Using clickstream data from 3873 individuals in the U.S., we find that this effect in fact exists only for online media piracy technologies that are perceived as legally ambiguous. Specifically, more conservative individuals, who typically have lower ambiguity intolerance, use (legal but ambiguously perceived) pirated streaming websites less, while there is no difference for the (clearly illegal) use of pirated file sharing websites.

Learning to Rank an Assortment of Products

We consider the product-ranking challenge that online retailers face when their customers typically behave as “window shoppers.” They form an impression of the assortment after browsing products ranked in the initial positions and then decide whether to continue browsing. We design online learning algorithms for product ranking that maximize the number of customers who engage with the site. Customers’ product preferences and attention spans are correlated and unknown to the retailer; furthermore, the retailer cannot exploit similarities across products, owing to the fact that the products are not necessarily characterized by a set of attributes. We develop a class of online learning-then-earning algorithms that prescribe a ranking to offer each customer, learning from preceding customers’ clickstream data to offer better rankings to subsequent customers. Our algorithms balance product popularity with diversity, the notion of appealing to a large variety of heterogeneous customers. We prove that our learning algorithms converge to a ranking that matches the best-known approximation factors for the offline, complete information setting. Finally, we partner with Wayfair — a multibillion-dollar home goods online retailer — to estimate the impact of our algorithms in practice via simulations using actual clickstream data, and we find that our algorithms yield a significant increase (5–30%) in the number of customers that engage with the site. This paper was accepted by J. George Shanthikumar for the Management Science Special Issue on Data-Driven Prescriptive Analytics.

On the Use of Generative Deep Learning Approaches for Generating Hidden Test Scripts

With the advent of web 2.0, web application architectures have been evolved, and their complexity has grown enormously. Due to the complexity, testing of web applications is getting time-consuming and intensive process. In today’s web applications, users can achieve the same goal by performing different actions. To ensure that the entire system is safe and robust, developers try to test all possible user action sequences in the testing phase. Since the space of all the possibilities is enormous, covering all user action sequences can be impossible. To automate the test script generation task and reduce the space of the possible user action sequences, we propose a novel method based on long short-term memory (LSTM) network for generating test scripts from user clickstream data. The experiment results clearly show that generated hidden test sequences are user-like sequences, and the process of generating test scripts with the proposed model is less time-consuming than writing them manually.

Personalized Online Education Learning Strategies Based on Transfer Learning Emotion Classification Model

Due to the epidemic, online course learning has become a major learning method for students worldwide. Analyzing its massive data from the massive online education platforms becomes a challenge because most learners watch online instructional videos. Thus, analyzing learners’ learning behaviors is beneficial to implement personalized online learning strategies with sentiment classification models. To this end, we propose a context-aware network model based on transfer learning that aims to predict learner performance by solving learners’ problems and improving the educational process, contributing to a comprehensive analysis of such student behavior and exploring various learning models in MOOC video interactions. In addition, we visualize and analyze MOOC video interactions, enabling course instructors and education professionals to analyze clickstream data generated by learners interacting with course videos. The experimental results show that, in the process of “massive data mining,” personalized learning strategies of this model can efficiently enhance students’ interest in learning and enable different types of students to develop personalized online education learning strategies.

Understanding student learning pathways in traditional online history courses: utilizing process mining analysis on clickstream data

PurposeThis study explores ongoing research into self-mapped learning pathways that students utilize to move through a course when given two modalities to choose from: one that is instructor-led and one that is student-directed.Design/methodology/approachProcess mining analysis was utilized to examine and cluster clickstream data from an online college-level History course designed with dual modality choices. This paper examines some of the results from different approaches to clustering the available data.FindingsBy examining how often students interacted with others, whether they were more internal or external facing with their pathway choices, and whether or not they completed a learning pathway, this study identified five general tactics from the data: Individualistic Internal; Non-completing Internal; Completing, Interactive Internal; Completing, Interactive, and Reflective and Completing External. Further analysis of when students used each tactic led to the identification of four different strategies that learners utilized during class sessions.Practical implicationsThe results of this analysis could potentially lead to the creation of customizable design models that can assist learners as they navigate modality choices in learner-centered or less-structured learning design methodologies.Originality/valueFew courses are designed to give the learners the options to follow the instructor or create their own learning pathway. Knowing how to identify what choices a learner might take in these scenarios is even less explored. Preliminary data for this paper was originally presented as a poster session at the Learning Analytics and Knowledge conference in 2019.

An Empirical Evaluation of Online Continuous Authentication and Anomaly Detection Using Mouse Clickstream Data Analysis

While the password-based authentication used in social networks, e-mail, e-commerce, and online banking is vulnerable to hackings, biometric-based continuous authentication systems have been used successfully to handle the rise in unauthorized accesses. In this study, an empirical evaluation of online continuous authentication (CA) and anomaly detection (AD) based on mouse clickstream data analysis is presented. This research started by gathering a set of online mouse-dynamics information from 20 participants by using software developed for collecting mouse information, extracting approximately 87 features from the raw dataset. In contrast to previous work, the efficiency of CA and AD was studied using different machine learning (ML) and deep learning (DL) algorithms, namely, decision tree classifier (DT), k-nearest neighbor classifier (KNN), random forest classifier (RF), and convolutional neural network classifier (CNN). User identification was determined by using three scenarios: Scenario A, a single mouse movement action; Scenario B, a single point-and-click action; and Scenario C, a set of mouse movement and point-and-click actions. The results show that each classifier is capable of distinguishing between an authentic user and a fraudulent user with a comparatively high degree of accuracy.