Future Learning Tasks Research Articles

Nursing students’ satisfaction with the instructional design of a computer-based virtual learning environment for mathematical medication learning

ABSTRACT Computer-based virtual learning environments (CBVLEs) have attracted attention as a learning innovation that can foster students’ self-efficacy and intrinsic motivation. Research on the instructional design regarding these aspects of learning in a virtual learning environment is rather piecemeal. This study investigates the instructional design of a CBVLE for mathematical medication learning by nursing students in vocational education. The instructional design was based on a task-centered approach, and students’ future learning tasks formed the backbone. We examine the extent to which the CBVLE fostered the nursing students’ mathematical learning, self-efficacy, and intrinsic motivation, and the ways in which the design components of the CBVLE met nursing students’ satisfaction. In total, 118 nursing students were assigned to four groups, with or without extra support from worked examples, and were trained via the CBVLE on mathematical medication learning tasks over four consecutive weeks. Students were pre- and post-tested on their mathematical medication learning, self-efficacy, and intrinsic motivation. Students also rated their satisfaction with the instructional design. Our results showed that the CBVLE fostered nursing students’ mathematical medication learning, self-efficacy, and intrinsic motivation, but no significant differences were found between the four conditions. Overall, student satisfaction was above average. The design components were able to predict nursing students’ mathematical medication learning, self-efficacy, and intrinsic motivation.

Rumors Detection Based on Lifelong Machine Learning

The amount of training data is small in the field of Weibo rumor detection, and online news changes constantly, but the existing models do not have the ability of continuous learning, which means they cannot achieve knowledge accumulation and update. Hence, they will require many training examples to improve their detection ability. In contrast, Lifelong Machine Learning (LML) paradigm has the capability of continuous learning, which retains the knowledge learned in the past and uses it to help future learning tasks. After learning, some knowledge is updated. With the growth and update of knowledge, the performance of each task will be better and better. Hence, we use this paradigm to build a Weibo rumor detection model. First, we extracted three types of features based on content, user, and propagation from Weibo events and proposed three new propagation features and Bidirectional Encoder Representations from Transformers (BERT) semantic features of source message for rumor detection. We then used Simulated Annealing (SA) to improve the Genetic Algorithm (GA), which was called GA-SA used to search for the best global minimum feature subset to enhance the classification effect of the Efficient Lifelong Learning Algorithm (ELLA) on rumors. The ELLA transfers knowledge to learn new tasks and refines knowledge over time to maximize performance across all tasks in the continuous learning process. The proposed model is called GA-SA-ELLA. The experimental results show that our model could achieve superior detection results even with less training data for each task.

The need to disentangle assessment and feedback in higher education

ABSTRACT In contemporary higher education systems, the processes of assessment and feedback are often seen as coexisting activities. As a result, they have become entangled in both policy and practice, resulting in a conceptual and practical blurring of their unique purposes. In this paper, we present a critical examination of the issues created by the entanglement of assessment and feedback, arguing that it is important to ensure that the legitimate purposes of both feedback and assessment are not compromised by inappropriate conflation of the two. We situate our argument in the shifting conceptual landscape of feedback, where there is increasing emphasis on students being active players in feedback processes working with and applying information from others to future learning tasks, rather than regarding feedback as a mechanism of transmission of information by teachers. We surface and critically discuss six problems created by the entanglement of assessment and feedback: students’ focus on grades; comments justifying grades rather than support learning; feedback too late to be useful; feedback subordinated to all other processes in course design; overemphasis on documentation of feedback; and the downgrading of feedback created by requirements for anonymous marking. We then propose a series of strategies for preserving the learning function of feedback, through models that give primacy to feedback within learning cycles. We conclude by offering suggestions for research and practice that seek to engage with the challenges created by the entanglement of assessment and feedback, and that maintain the unique purposes of assessment and feedback.

DeGAN: Data-Enriching GAN for Retrieving Representative Samples from a Trained Classifier

In this era of digital information explosion, an abundance of data from numerous modalities is being generated as well as archived everyday. However, most problems associated with training Deep Neural Networks still revolve around lack of data that is rich enough for a given task. Data is required not only for training an initial model, but also for future learning tasks such as Model Compression and Incremental Learning. A diverse dataset may be used for training an initial model, but it may not be feasible to store it throughout the product life cycle due to data privacy issues or memory constraints. We propose to bridge the gap between the abundance of available data and lack of relevant data, for the future learning tasks of a given trained network. We use the available data, that may be an imbalanced subset of the original training dataset, or a related domain dataset, to retrieve representative samples from a trained classifier, using a novel Data-enriching GAN (DeGAN) framework. We demonstrate that data from a related domain can be leveraged to achieve state-of-the-art performance for the tasks of Data-free Knowledge Distillation and Incremental Learning on benchmark datasets. We further demonstrate that our proposed framework can enrich any data, even from unrelated domains, to make it more useful for the future learning tasks of a given network.

Improving Named Entity Recognition in Vietnamese Texts by a Character-Level Deep Lifelong Learning Model

Named entity recognition (NER) is a fundamental task which affects the performance of its dependent task, e.g. machine translation. Lifelong machine learning (LML) is a continuous learning process, in which the knowledge base accumulated from previous tasks will be used to improve future learning tasks having few samples. Since there are a few studies on LML based on deep neural networks for NER, especially in Vietnamese, we propose a lifelong learning model based on deep learning with a CRFs layer, named DeepLML–NER, for NER in Vietnamese texts. DeepLML–NER includes an algorithm to extract the knowledge of “prefix-features” of named entities in previous domains. Then the model uses the knowledge in the knowledge base to solve the current NER task. Preprocessing and model parameter tuning are also investigated to improve the performance. The effect of the model was demonstrated by in-domain and cross-domain experiments, achieving promising results.

The Impact of Guided Metacognitive Feedback on Novice Programmers Using Learning by Teaching Environment

Learning-by-teaching is a powerful approach that enhances students to think deeply, orally and repeatedly. Several computer-based systems have been implemented where students play the teacher role and virtual agents play the tutee role. The existing systems focus on various domains, but none of them has considered programming problem solving. Additionally, the majority of these systems did not provide metacognitive support. They only focus on providing feedback as correct answers, and this type of feedback is called knowledge of correct response. However, this paper explores the influence of guided metacognitive feedback on novice programmers in a teachable agent environment. For that, a computer-based learning environment is built to enable the novice programmers to teach programming problem solving to an animated agent. It combines learning-by-teaching technique and metacognitive support in order to assist those beginners to acquire comprehensive learning on how to solve unfamiliar problems and prepare those programmers for future learning tasks. We conduct an experiment to compare the effect of the aforementioned feedbacks on the novice programmers’ performance in learning-byteaching paradigm. The results show that the metacognitive feedback has positive effect on novice programmers’ achievement of solving problems. In addition, providing metacognitive feedback as explicit feedback in learning-by-teaching paradigm improves the novices' abilities to estimate what they know and what they do not know about how to solve new programming problems.

Active Task Selection for Lifelong Machine Learning

In a lifelong learning framework, an agent acquires knowledge incrementally over consecutive learning tasks, continually building upon its experience. Recent lifelong learning algorithms have achieved nearly identical performance to batch multi-task learning methods while reducing learning time by three orders of magnitude. In this paper, we further improve the scalability of lifelong learning by developing curriculum selection methods that enable an agent to actively select the next task to learn in order to maximize performance on future learning tasks. We demonstrate that active task selection is highly reliable and effective, allowing an agent to learn high performance models using up to 50% fewer tasks than when the agent has no control over the task order. We also explore a variant of transfer learning in the lifelong learning setting in which the agent can focus knowledge acquisition toward a particular target task.