Generative artificial intelligence (GenAI) is increasingly embedded in higher education, yet little is known about how it shapes students’ cognitive engagement in authentic learning tasks. This study applies a learning analytics perspective to investigate the types and dynamics of cognition that emerge when students interact with GenAI during a question–answering task. We collected 294 GPT-4 assisted conversations comprising 618 student questions using a purpose-built system designed to support student GenAI questioning. Student–GenAI interactions were coded with Bloom’s revised taxonomy to capture levels of cognition, and then analyzed with two complementary learning analytics approaches: Epistemic Network Analysis (ENA), which identifies co-occurrence structures among cognitive categories, and Frequency-based Transition Network Analysis (FTNA), which models their sequential flow over time. The analyses revealed that prompts generated by the AI tool elicited higher cognitive processes, particularly Analyze with a spillover to Apply, whereas student-authored prompts clustered around Understand and Evaluate. FTNA showed that Remember commonly served as the entry point and Analyze as the main convergence node; although rare, Create occupied a central position when it appeared. Students often engaged in cyclical loops across Bloom’s categories rather than following a linear progression through the taxonomy. This work contributes to learning analytics by demonstrating how combining Bloom’s revised taxonomy with ENA and FTNA provides methodological innovations for capturing both the structural and temporal dynamics of cognition in GenAI-supported learning environments. Pedagogically, the findings highlight how learning analytics can inform the design of AI-generated and student-authored prompts to foster deeper cognitive engagement, thus offering actionable insights for empowering students and guiding educators in higher education contexts.

Poetry represents the oldest and most esteemed literary form, allowing poets to convey ideas while carefully attending to elements such as meaning, coherence, poetic quality, and fluency. Notably, the creation of good poetry entails considerations of rhyme and meter. With the advent of artificial intelligence (AI), significant advancements have been made in automatic text generation, primarily within languages such as English and Chinese. However, the generation of Urdu poetry presents a unique challenge due to the language’s inherent ambiguity, cultural and historical nuances, and the demand for creativity. The existing body of literature has only marginally explored Urdu prose and has almost entirely overlooked the domain of Urdu poetry generation, primarily due to the scarcity of comprehensive training data. In response to this deficiency, this research endeavor addresses this challenge. It begins by introducing a specialized Urdu poetry dataset adhering to a specific meter, “behr-e-khafeef,” which incorporates approximately 20,000 couplets from the Rekhta repository. Subsequently, a character-based encoding methodology is proposed to transform these couplets into a numerical representation, assigning a distinct identifier to each character. The generation process initiates with the creation of the first verse through a character-level LSTM, followed by the application of a machine translation technique, specifically sequence-to-sequence learning, to formulate the second verse based on the first. The generated poetry is subjected to evaluation based on metrics, including BLEU scores. Additionally, an expert panel of Urdu poets is engaged to conduct a human assessment of the generated couplets, with the evaluation encompassing critical dimensions such as meaning, coherence, poetic quality, and fluency. Our findings are juxtaposed with existing poetry generation systems, demonstrating a notable advancement in the state-of-the-art, as evidenced by a BLEU score of 0.23. The research culminates with the presentation of prospective avenues for further exploration, aimed at inspiring the scholarly community to enhance the domain of poetry generation and augment existing contributions in this field.

To meet the growing demand for future professionals who can present information to an audience and create quality written products, educators are increasingly assigning writing assignments that require students to gather information from multiple sources, reorganise and reinterpret knowledge from source materials, and plan for rhetorical structure goals in order to meet the task requirements. When evaluating an essay coherence, scorers manually look for the presence of required rhetorical categories, which takes time. Supervised Machine Learning (ML) techniques have proven to be an effective tool for automatic detection of rhetorical categories that approximate students’ cognitive engagement with source information. Previous studies that addressed this problem used relatively small datasets and reported relatively low kappa scores for accuracy, limiting the use of such models in real-world scenarios. Moreover, to empower educators to effectively evaluate the overall quality of students’ writing, the associations between the sequential patterns of rhetorical categories in students’ writing and writing performance must be examined, which remains largely unexplored in educational domain. Therefore, to fill these gaps, our study aimed to i) investigate the impact of data augmentation approaches on the performance of deep learning algorithms in classifying rhetorical categories in student essays according to Bloom‘s taxonomy ii) and explore the sequential patterns of rhetorical categories in students’ writing that can influence writing performance. Our findings showed that deep learning-based model BERT on Easy Data Augmentation (EDA) based augmented data achieved 20% higher Cohen’s kappa than normal (non-augmented) data, and we discovered that students in different performance groups were statistically different in terms of rhetorical patterns. Our proposed study is valuable in terms of building a data analytic foundation that can be used to create formative feedback on students’ writings based on the patterns of rhetorical categories to improve essay quality.

Graph representation methods have recently become the de facto standard for downstream machine learning tasks on graph-structured data and have found numerous applications, e.g., drug discovery & development, recommendation, and forecasting. However, the existing methods are specially designed to work in a centralized environment, which limits their applicability to small or medium-sized graphs. In this work, we present a graph embedding method that extracts graph representations in a distributed environment with independent and parallel machines. The proposed method is built-upon the existing approach, distributed graph statistical distance (DGSD), to enhance the scalability on large graphs. The key innovation of our work lies in the proposition of a batching mechanism for client-server message passing, which reduces communication overhead during the computation of the distance matrix. In addition, we present a sampling approach for computing pairwise distances between the nodes to compute the desired graph embedding. Moreover, we systematically explore six distinct variations of a distributed graph embeddings and subsequently subject them to comprehensive evaluation. Our extensive evaluations on over 20 graph datasets and ten baseline methods demonstrate improved running time and comparative classification accuracy compared to state-of-the-art embedding techniques.

Essay writing has become one of the most common learning tasks assigned to students enrolled in various courses at different educational levels, owing to the growing demand for future professionals to effectively communicate information to an audience and develop a written product (i.e. essay). Evaluating a written product requires scorers who manually examine the existence of rhetorical categories, which is a time-consuming task. Machine Learning (ML) approaches have the potential to alleviate this challenge. As a result, several attempts have been made in the literature to automate the identification of rhetorical categories using Rhetorical Structure Theory (RST). However, RST do not provide information regarding students’ cognitive level, which motivates the use of Bloom’s Taxonomy. Therefore, in this research we propose to: i) investigate the extent to which classification of rhetorical categories can be automated based on Bloom’s taxonomy by comparing the traditional ML classifiers with the pre-trained language model BERT, ii) explore the associations between rhetorical categories and writing performance. Our results showed that BERT model outperformed the traditional ML-based classifiers with 18% better accuracy, indicating it can be used in future analytics tool. Moreover, we found a statistical difference between the associations of rhetorical categories in low-achiever, medium-achiever and high-achiever groups which implies that rhetorical categories can be predictive of writing performance.

Abstract The quality of scientific publications can be measured by quantitative indices such as the h‐index, Source Normalized Impact per Paper, or g‐index. However, these measures lack to explain the function or reasons for citations and the context of citations from citing publication to cited publication. We argue that citation context may be considered while calculating the impact of research work. However, mining citation context from unstructured full‐text publications is a challenging task. In this paper, we compiled a data set comprising 9,518 citations context. We developed a deep learning‐based architecture for citation context classification. Unlike feature‐based state‐of‐the‐art models, our proposed focal‐loss and class‐weight‐aware BiLSTM model with pretrained GloVe embedding vectors use citation context as input to outperform them in multiclass citation context classification tasks. Our model improves on the baseline state‐of‐the‐art by achieving an F1 score of 0.80 with an accuracy of 0.81 for citation context classification. Moreover, we delve into the effects of using different word embeddings on the performance of the classification model and draw a comparison between fastText, GloVe, and spaCy pretrained word embeddings.

This research aimed to explore the relationship between different indicators of the depth and quality of participation in computer-mediated learning environments. By using network analyses and statistical tests, we discovered significant associations between the cognitive presence phases of the Community of Inquiry framework and speech acts, and examined the impact of two different instructional interventions on these associations. We found that there are strong associations between some speech acts and cognitive presence phases. In addition, the study revealed that the association between speech acts and cognitive presence is moderated by external facilitation, but not affected by user role assignment. The results suggest that speech acts can plausibly be used to provide feedback in relation to cognitive presence and can potentially be used to increase the generalizability of cognitive presence classification.

Student retention is a widely recognized challenge in the educational community to assist the institutes in the formation of appropriate and effective pedagogical interventions. This study intends to predict the students at-risk of low performances during an on-going course, those at-risk of graduating late than the tentative timeline and predicting the capacity of students in a campus. The data constitutes of demographics, learning, academic and educational related attributes which are suitable to deploy various machine learning algorithms for the prediction of at-risk students. For class balancing, Synthetic Minority Over Sampling Technique, is also applied to eliminate the imbalance in the academic award-gap performances and late/timely graduates. Results reveal the effectiveness of the deployed techniques with Long short-term Memory (LSTM) outperforming other models for early prediction of at-risk students. The main contribution of this work is a machine learning approach capable of enhancing the academic decision making related to student performance.

Abstract Most existing studies are focused on popular languages like English, Spanish, Chinese, Japanese, and others, however, limited attention has been paid to Urdu despite having more than 60 million native speakers. In this paper, we develop a deep learning model for the sentiments expressed in this under‐resourced language. We develop an open‐source corpus of 10,008 reviews from 566 online threads on the topics of sports, food, software, politics, and entertainment. The objectives of this work are bi‐fold (a) the creation of a human‐annotated corpus for the research of sentiment analysis in Urdu; and (b) measurement of up‐to‐date model performance using a corpus. For their assessment, we performed binary and ternary classification studies utilizing another model, namely long short‐term memory (LSTM), recurrent convolutional neural network (RCNN) Rule‐Based, N‐gram, support vector machine , convolutional neural network, and LSTM. The RCNN model surpasses standard models with 84.98% accuracy for binary classification and 68.56% accuracy for ternary classification. To facilitate other researchers working in the same domain, we have open‐sourced the corpus and code developed for this research.

Recommender systems use machine-learning techniques to make predictions about resources. The medical field is one where much research is currently being conducted on recommender system utility. In the last few years, the amount of information available online that relates to healthcare has increased tremendously. Patients nowadays are more aware and look for answers to healthcare problems online. This has resulted in a dire need of an effective reliable online system to recommend the physician that is best suited to a particular patient in a limited time. In this article, a hybrid doctor-recommender system is proposed, by combining different recommendation approaches: content base, collaborative and demographic filtering to effectively tackle the issue of doctor recommendation. The proposed system addresses the issue of personalization through analysing patient's interest towards selecting a doctor. It uses a novel adoptive algorithm to construct a doctor's ranking function. Moreover, this ranking function is used to translate patients’ criteria for selecting a doctor into a numerical base rating, which will eventually be used in the recommendation of doctors. The system has been evaluated thoroughly, and result show that recommendations are reasonable and can fulfil patient's demand for reliable doctor's selection effectively.

The current evolution in multidisciplinary learning analytics research poses significant challenges for the exploitation of behavior analysis by fusing data streams toward advanced decision-making. The identification of students that are at risk of withdrawals in higher education is connected to numerous educational policies, to enhance their competencies and skills through timely interventions by academia. Predicting student performance is a vital decision-making problem including data from various environment modules that can be fused into a homogenous vector to ascertain decision-making. This research study exploits a temporal sequential classification problem to predict early withdrawal of students, by tapping the power of actionable smart data in the form of students' interactional activities with the online educational system, using the freely available Open University Learning Analytics data set by employing deep long short-term memory (LSTM) model. The deployed LSTM model outperforms baseline logistic regression and artificial neural networks by 10.31% and 6.48% respectively with 97.25% learning accuracy, 92.79% precision, and 85.92% recall.

In the present era of Big Data, with continuously increasing amounts of user-generated content, it is becoming a challenge to understand the relation between the content that is available on the Web and the users who are generating that content. Researchers have come up with many ways to understand today's Web better. One of the recently introduced concepts is a Web observatory (WO). This article provides a deep understanding about web observatories. It discusses the status of existing WO systems. The article investigates and gathers the common practices of WOs. This research has implications for researchers and communities in the adoption of the WO concept. The article highlights the challenges of WOs, such as data crawling, privacy and security. It also provides future research and development directions. The article provides a comparative analysis of existing WOs. It discusses the architecture of WOs. It presents components of a WO in a coherent manner and finally provides insights into challenges and limitations of WOs.

In our days, the interaction of Behaviour and Information Technology is challenged by emerging technologies, including cognitive computing and business intelligence. The learning domain has many fe...

Learning analytics is an emerging field of research, motivated by the wide spectrum of the available educational information that can be analysed to provide a data-driven decision about various learning problems. This study intends to examine the research landscape of learning analytics to deliver a comprehensive understanding of the research activities in this multidisciplinary field, using scientific literature from the Scopus database. An array of state-of-the-art bibliometric indices is deployed on 2811 procured publication datasets: publication counts, citation counts, co-authorship patterns, citation networks and term co-occurrence. The results indicate that the field of learning analytics appears to have been instantiated around 2011; thus, before this time period no significant research activity can be observed. The temporal evolution indicates that the terms ‘students’, ‘teachers’, ‘higher education institutions’ and ‘learning process’ appear to be the major components of the field. More recent trends in the field are the tools that tap into Big Data analytics and data mining techniques for more rational data-driven decision-making services. A future direction research depicts a need to integrate learning analytics research with multidisciplinary smart education and smart library services. The vision towards smart city research requires a meta-level of smart learning analytics value integration and policy-making.

In this article we present a new semantic and syntactic-based method for external plagiarism detection. In the proposed approach, latent dirichlet allocation (LDA) and parts of speech (POS) tags are used together to detect plagiarism between the sample and a number of source documents. The basic hypothesis is that considering semantic and syntactic information between two text documents may improve the performance of the plagiarism detection task. Our method is based on two steps, naming, which is a pre-processing where we detect the topics from the sentences in documents using the LDA and convert each sentence in POS tags array; then a post processing step where the suspicious cases are verified purely on the basis of semantic rules. For two types of external plagiarism (copy and random obfuscation), we empirically compare our approach to the state-of-the-art N-gram based and stop-word N-gram based methods and observe significant improvements.

Rapid advancements in technology over the last two decades played a key role in increasing the utilizations of mobile banking applications. Banks are providing this service to allow their customers to access their accounts anytime anywhere. Although many studies have assessed user satisfaction of mobile banking, most of these studies have been done in developed countries and only a few have compared mobile banking in developing with developed countries. This study assessed user satisfaction of mobile banking in both the United Kingdom and Saudi Arabia. Above 100 online questionnaires were collected from individuals who have experience in using mobile banking applications in both the UK and Saudi Arabia. The results show that system quality has a significant effect on customer satisfaction in the UK, but not in Saudi Arabia. Secondly, both information quality and interface design quality have significant effects on customer satisfaction in both the UK and Saudi Arabia. However, the overall findings from this study suggest that respondents are more satisfied with mobile banking in the UK than with mobile banking in Saudi Arabia.

Nowadays, wireless body area networks (WBANs) systems have adopted cloud computing (CC) technology to overcome limitations such as power, storage, scalability, management, and computing. This amalgamation of WBANs systems and CC technology, as sensor‐cloud infrastructure (S‐CI), is aiding the healthcare domain through real‐time monitoring of patients and the early diagnosis of diseases. Hence, the distributed environment of S‐CI presents new threats to patient data privacy and security. In this paper, we review the techniques for patient data privacy and security in S‐CI. Existing techniques are classified as multibiometric key generation, pairwise key establishment, hash function, attribute‐based encryption, chaotic maps, hybrid encryption, Number Theory Research Unit, Tri‐Mode Algorithm, Dynamic Probability Packet Marking, and Priority‐Based Data Forwarding techniques, according to their application areas. Their pros and cons are presented in chronological order. We also provide our six‐step generic framework for patient physiological parameters (PPPs) privacy and security in S‐CI: (1) selecting the preliminaries; (2) selecting the system entities; (3) selecting the technique; (4) accessing PPPs; (5) analysing the security; and (6) estimating performance. Meanwhile, we identify and discuss PPPs utilized as datasets and provide the performance evolution of this research area. Finally, we conclude with the open challenges and future directions for this flourishing research area.

Producing the Linked Open Data (LOD) is getting potential to publish high-quality interlinked data. Publishing such data facilitates intelligent searching from the Web of data. In the context of scientific publications, data about millions of scientific documents published by hundreds and thousands of publishers is in silence as it is not published as open data and ultimately is not linked to other datasets. In this paper the authors present SPedia: a semantically enriched knowledge base of data about scientific documents. SPedia knowledge base provides information on more than nine million scientific documents, consisting of more than three hundred million RDF triples. These extracted datasets, allow users to put sophisticated queries by employing semantic Web techniques instead of relying on keyword-based searches. This paper also shows the quality of extracted data by performing sample queries through SPedia SPARQL Endpoint and analyzing results. Finally, the authors describe that how SPedia can serve as central hub for the cloud of LOD of scientific publications.

Social media has revolutionized human communication and styles of interaction. Due to its effectiveness and ease, people have started using it increasingly to share and exchange information, carry out discussions on various events, and express their opinions. Various communities may have diverse sentiments about events and it is an interesting research problem to understand the sentiments of a particular community for a specific event. In this article, the authors propose a framework CommuniMents which enables us to identify the members of a community and measure the sentiments of the community for a particular event. CommuniMents uses automated snowball sampling to identify the members of a community, then fetches their published contents (specifically tweets), pre-processes the contents and measures the sentiments of the community. The authors perform qualitative and quantitative evaluation for a variety of real world events to validate the effectiveness of the proposed framework.

With the convergence of pervasive mobile computing and social networking, interest has grown significantly in software systems and applications that are aware of users' social context to make pervasive applications more intelligent and accessible. Thus, socially-aware systems have further advanced context-aware systems taking account of human social context such as social relationships to enable the attainment of users' tasks in different domains. However, social context-awareness introduces a variety of software engineering challenges. In this paper, we address these challenges by proposing a software engineering process that provides a methodological framework for developing various types of socially-aware applications from requirements elicitation through to concrete implementation. We provide context models and software infrastructure to assist developers in rapid prototyping. We also present two case studies to demonstrate the feasibility and applicability of our software engineering process by presenting how this process can be used to develop two different types of socially-aware applications utilizing our model and infrastructure. Finally, we evaluate our software engineering approach with respect to a set of software quality metrics.

It is of high relevance to the LAK community to explore blended learning scenarios where students can interact at diverse digital and physical learning spaces. This workshop aims to gather the sub-community of LAK researchers, learning scientists and researchers from other communities, interested in ubiquitous, mobile and/or face-to-face learning analytics. An overarching concern is how to integrate and coordinate learning analytics to provide continued support to learning across digital and physical spaces. The goals of the workshop are to share approaches and identify a set of guidelines to design and connect Learning Analytics solutions according to the pedagogical needs and contextual constraints to provide support across digital and physical learning spaces.

Developing large-scale exams for a massive number of preparatory year students by a small designated committee of exam developers is a common practice in colleges and universities around the world. However, in adopting this approach, some higher institutions might be confronted with some obstacles such as the unavailability of skilful exam developers and proper resources. To address this issue, we proposed CrowdyQ, a virtual crowdsourcing platform, to investigate the impact of involving a large number of same-subject university instructors in the creation and evaluation of question items in a collaborative manner for the purpose of developing high quality exams. This paper describes the development of CrowdyQ platform and it discusses briefly its potential significance in developing high quality large-scale exam questions in a university context.

Twitter is a popular micro-blogging service for sharing short messages called tweets. Tweets provide public opinion on various topics. Currently twitter presents search results in form of a flat list, sorted either by popularity or by recency. These search results limit the possibility of identifying diverse latent topics covered by the tweets. One way to better understand the tweets is to cluster them where each cluster depicts a latent topic. Suitable clustering algorithms are required to cluster streaming data and map new data into existing clusters. To address this, we propose in this paper a framework called INTWEEMS (INcremental clustering of TWEEt streaMS) which clusters tweets in real-time, adjusts new tweets into existing clusters (incrementally), and provides visualization of clusters that helps in identifying latent topics and sub-topics within the tweets. This paper describes the INTWEEMS framework and its implementation.

As mobile devices are getting more powerful and more affordable the use of online educational multimedia is also getting very prevalent. Limited battery power is nevertheless a major restricting factor as streaming multimedia drains battery power quickly. Many battery efficient multimedia adaptation techniques have been proposed that achieve battery efficiency by lowering presentation quality of entire multimedia. Adaptation is usually done without considering any impact on the information contents of multimedia. In this paper, based on the results of an experimental study, we argue that without considering any negative impact on information contents of multimedia the adaptation may negatively impact the learning process. Some portions of the multimedia that require a higher visual quality for conveying learning information may lose their learning effectiveness in the adapted lowered quality. We report results of our experimental study that indicate that different parts of the same learning multimedia do not have same minimum acceptable quality. This strengthens the position that power-saving adaptation techniques for educational multimedia must be developed that lower the quality of multimedia based on the needs of its individual fragments for successfully conveying learning information.

Using online multimedia content on mobile devices is a power hungry activity and drains battery power very quickly. This poses a big challenge in using mobile devices with limited battery power for learning purposes using online educational multimedia. Multimedia adaptation techniques have been developed that preserve battery power by lowering multimedia quality. These adaptation techniques do not provide users with any power-saving options and the adaptation is done automatically without involvement of users. In this paper, we propose a Learner-Battery Interaction model that suggests involving learners in the adaptation process. The idea is to provide learners with power-saving options and relevant feedback about the form of adapted multimedia in advance. This will help leaners in making informed power-saving decisions for adaptation. We implemented the model in a prototype system and conducted an evaluation in the form of a user study.

As a result of tremendous enhancements in the capabilities of mobile devices and availability of higher data rate mobile internet, the use of online multimedia learning resources on mobile devices is increasingly becoming popular. Limited Battery Power of mobile devices, however, is still one big challenge in Mobile Learning. High Quality multimedia learning resources are power hungry and if used on mobile devices drain battery power rapidly limiting learning opportunities on the move. Lack of significant improvements in battery capacities has resulted in significant interest in battery power saving techniques. Existing power-saving streaming multimedia adaptation techniques tend to extend battery life by reducing quality of multimedia making them susceptible to information loss. This loss may affect the learning content efficacy and jeopardizes the learning process. To the best of our knowledge, no previous work has considered the learning content efficacy in multimedia streaming adaptation mechanism. In this paper, we present MoBELearn system, which is a prototype implementation of our proposed Content Aware Power Saving Educational Multimedia Adaptation (CAPS-EMA) approach. We demonstrate battery efficiency in educational multimedia streaming while keeping the adapted resource suitable for learning. We also describe our semantic metamodel for educational multimedia resource that support our energy efficient adaptation technique.