Search results

This chapter describes a successful research-developed representation construction approach to teaching and learning that links student learning and engagement with the epistemic practices of science. This approach involves challenging students to generate and negotiate the representations (text, graphs, models, diagrams) that constitute the discursive practices of science, rather than focusing on the text-based, definitional versions of concepts. The representation construction approach is based on sequences of representational challenges that involve students constructing representations to actively explore and make claims about phenomena. The key principles of the representation construction approach, considered a form of directed inquiry, are outlined with illustrations from case studies of whole topics in forces and astronomy within several middle-years’ science classrooms. This chapter also outlines the manner in which the representation construction approach has been translated into wider scale implementation through a large-scale Professional Development (PD) workshop program. Issues associated with wider scale implementation of the approach are discussed.

The purpose of this paper is to describe a new approach to sentence representation learning leading to text classification using Bidirectional Encoder Representations from Transformers (BERT) embeddings. This work proposes a novel BERT-convolutional neural network (CNN)-based model for sentence representation learning and text classification. The proposed model can be used by industries that work in the area of classification of similarity scores between the texts and sentiments and opinion analysis.

The approach developed is based on the use of the BERT model to provide distinct features from its transformer encoder layers to the CNNs to achieve multi-layer feature fusion. To achieve multi-layer feature fusion, the distinct feature vectors of the last three layers of the BERT are passed to three separate CNN layers to generate a rich feature representation that can be used for extracting the keywords in the sentences. For sentence representation learning and text classification, the proposed model is trained and tested on the Stanford Sentiment Treebank-2 (SST-2) data set for sentiment analysis and the Quora Question Pair (QQP) data set for sentence classification. To obtain benchmark results, a selective training approach has been applied with the proposed model.

On the SST-2 data set, the proposed model achieved an accuracy of 92.90%, whereas, on the QQP data set, it achieved an accuracy of 91.51%. For other evaluation metrics such as precision, recall and F1 Score, the results obtained are overwhelming. The results with the proposed model are 1.17%–1.2% better as compared to the original BERT model on the SST-2 and QQP data sets.

The novelty of the proposed model lies in the multi-layer feature fusion between the last three layers of the BERT model with CNN layers and the selective training approach based on gated pruning to achieve benchmark results.

The purpose of this paper is to propose a game-based learning (GBL) content design model that replicates the two-dimensional Bloom cognitive process in GBL units. The proposed model, called the knowledge and cognitive-process representation (KCR) model, enables a game player to access three types of Bloom knowledge by allowing the learner to experience-related cognitive processes that can be replicated in the GBL units via appropriate representation approaches.

To validate the feasibility of the proposed KCR model, 14 GBL units for a Cisco-certified network associate (CCNA) certification training program were designed and installed on several servers. Players played the GBL units via internet browsers. According to the problem-solving theory, three game components, including a tool, feedback, and goal, are necessary for game playing and should be adopted to implement three sub-cognitive processes. A three-phase experiment was performed for one year. Subjects were university sophomores and a randomized block experiment design was implemented.

The experimental results show that, compared with a traditional web-based learning platform, the GBL platform is more efficient and it enables learners to achieve improved learning performance. In addition, most hypotheses support the fact that particular cognizance processes should be implemented by a specific representation approach in GBL. Finally, a KCR model for GBL content design is inferred to represent a cognitive process appropriately that can be referenced for both the digital content instructor and the game developer.

Because the CCNA training material does not include meta-knowledge of Bloom knowledge type and the creation of the Bloom cognitive process, the KCR model should be further extended. In addition, others certification training materials (such as Oracle DBA, Java programmer) can be implemented on the basis of the KCR model for general validation as further research.

Players can acquire specific types of knowledge, such as factual knowledge, by experiencing a particular cognitive process, such as the “remembering & understanding” processes, which can be represented with a computer tool. The KCR model can provide both the instructor and the game developer with design recommendations and accelerate GBL content implementation.

GBL is a learning platform that can stimulate a learner by improving the motivation to learn and the learning experience. To ensure high-learning performance, the learner should perform specific cognitive processes and acquire knowledge. This research proposes a content design model for GBL units that appropriately replicate the Bloom framework in a computer game.

The purpose of this paper is to review recent developments in educational informatics relating to the provision by information systems of pedagogical support to web‐based learners, and to propose further investigation of the feasibility and potential value of web‐based “conversational” information systems to complement adaptive hypermedia and information retrieval systems.

The potential of Pask's conversation theory is considered as a potentially useful framework for the development of information systems capable of providing pedagogical support for web‐based learners, complementary to that provided by existing computer‐assisted learning and adaptive hypermedia systems. The potential role and application of entailment meshes are reviewed in relation to other forms of knowledge representation including classifications, semantic networks, ontologies and representations based on knowledge space theory.

Concludes that conversation theory could be a useful framework to support the development of web‐based “conversational” information that would complement aspects of computer‐assisted learning, adaptive hypermedia and information retrieval systems. The entailment mesh knowledge representation associated with conversation theory provides the potential for providing particularly rich pedagogical support by virtue of its properties of cyclicity, consistency and connectivity, designed to support deep and enduring levels of understanding.

Although based on a considerable body of theoretical and empirical work relating to conversation theory, the paper remains speculative in that the gap is still great between, on the one hand, what has so far been achieved and, on the other, the practical realisation of its potential to enhance web‐based learning. Much work remains to be done in terms of exploring the extent to which procedures developed and benefits found in relatively small‐scale experimental contexts can effectively be scaled to yield enhanced support for “real world” learning‐related information behaviour.

The ideas of Pask, discussed in this paper, are capable of guiding the structuring of information according to parameters designed to facilitate deep and enduring understanding via interactive “conversational” engagement between the conceptual structures of information source authors and learners. If one can scale Pask's work to “real world” learning‐related information behaviour, one can increase the range of web‐based information systems and services capable of providing pedagogical support to web‐based learners.

Compared with traditional methods relying on manual teaching or system modeling, data-driven learning methods, such as deep reinforcement learning and imitation learning, show more promising potential to cope with the challenges brought by increasingly complex tasks and environments, which have become the hot research topic in the field of robot skill learning. However, the contradiction between the difficulty of collecting robot–environment interaction data and the low data efficiency causes all these methods to face a serious data dilemma, which has become one of the key issues restricting their development. Therefore, this paper aims to comprehensively sort out and analyze the cause and solutions for the data dilemma in robot skill learning.

First, this review analyzes the causes of the data dilemma based on the classification and comparison of data-driven methods for robot skill learning; Then, the existing methods used to solve the data dilemma are introduced in detail. Finally, this review discusses the remaining open challenges and promising research topics for solving the data dilemma in the future.

This review shows that simulation–reality combination, state representation learning and knowledge sharing are crucial for overcoming the data dilemma of robot skill learning.

To the best of the authors’ knowledge, there are no surveys that systematically and comprehensively sort out and analyze the data dilemma in robot skill learning in the existing literature. It is hoped that this review can be helpful to better address the data dilemma in robot skill learning in the future.

Matching instances of the same entity, a task known as entity resolution, is a key step in the process of data integration. This paper aims to propose a deep learning network that learns different representations of Web entities for entity resolution.

To match Web entities, the proposed network learns the following representations of entities: embeddings, which are vector representations of the words in the entities in a low-dimensional space; convolutional vectors from a convolutional layer, which capture short-distance patterns in word sequences in the entities; and bag-of-word vectors, created by a bow layer that learns weights for words in the vocabulary based on the task at hand. Given a pair of entities, the similarity between their learned representations is used as a feature to a binary classifier that identifies a possible match. In addition to those features, the classifier also uses a modification of inverse document frequency for pairs, which identifies discriminative words in pairs of entities.

The proposed approach was evaluated in two commercial and two academic entity resolution benchmarking data sets. The results have shown that the proposed strategy outperforms previous approaches in the commercial data sets, which are more challenging, and have similar results to its competitors in the academic data sets.

No previous work has used a single deep learning framework to learn different representations of Web entities for entity resolution.

Student competency in science learning relies on students being able to interpret and use multimodal representations to communicate understandings. Moreover, collaborative learning, in which students may share physiological arousal, can positively affect group performance. This paper aims to observe changes in student attitudes and beliefs, physiology (electrodermal activity; EDA) and content knowledge before and after a multimodal, cooperative inquiry, science teaching intervention to determine associations with productive science learning and increased science knowledge.

A total of 214 students with a mean age of 11 years 6 months from seven primary schools participated in a multimodal, cooperative inquiry, science teaching intervention for eight weeks during a science curriculum unit. Students completed a series of questionnaires pertaining to attitudes and beliefs about science learning and science knowledge before (Time 1) and after (Time 2) the teaching intervention. Empatica E3 wristbands were worn by students during 1 to 3 of their regularly scheduled class sessions both before and after the intervention.

Increases in EDA, science knowledge, self-efficacy and a growth mindset, and decreases in self-esteem, confidence, motivation and use of cognitive strategies, were recorded post-intervention for the cohort. EDA was positively correlated with science knowledge, but negatively correlated with self-efficacy, motivation and use of cognitive strategies. Cluster analysis suggested three main clusters of students with differing physiological and psychological profiles.

First, teachers need to be aware of the importance of helping students to consolidate their current learning strategies as they transition to new learning approaches to counter decreased confidence. Second, teachers need to know that an effective teaching multimodal science intervention can not only be associated with increases in science knowledge but also increases in self-efficacy and movement towards a growth mindset. Finally, while there is evidence that there are positive associations between physiological arousal and science knowledge, physiological arousal was also associated with reductions in self-efficacy, intrinsic motivation and the use of cognitive strategies. This mixed result warrants further investigation.

Overall, this study proposes a need for teachers to counter decreased confidence in students who are learning new strategies, with further research required on the utility of monitoring physiological markers.

Session-based recommendation aims to predict the user's next preference based on the user's recent activities. Although most existing studies consider the global characteristics of items, they only learn the global characteristics of items based on a single connection relationship, which cannot fully capture the complex transformation relationship between items. We believe that multiple relationships between items in learning sessions can improve the performance of session recommendation tasks and the scalability of recommendation models. At the same time, high-quality global features of the item help to explore the potential common preferences of users.

This work proposes a session-based recommendation method with a multi-relation global context–enhanced network to capture this global transition relationship. Specifically, we construct a multi-relation global item graph based on a group of sessions, use a graded attention mechanism to learn different types of connection relations independently and obtain the global feature of the item according to the multi-relation weight.

We did related experiments on three benchmark datasets. The experimental results show that our proposed model is superior to the existing state-of-the-art methods, which verifies the effectiveness of our model.

First, we construct a multi-relation global item graph to learn the complex transition relations of the global context of the item and effectively mine the potential association of items between different sessions. Second, our model effectively improves the scalability of the model by obtaining high-quality item global features and enables some previously unconsidered items to make it onto the candidate list.

Drug repurposing involves the identification of new applications for existing drugs. Owing to the enormous rise in the costs of pharmaceutical R&D, several pharmaceutical companies are leveraging repurposing strategies. Parkinson's disease is the second most common neurodegenerative disorder worldwide, affecting approximately 1–2 percent of the human population older than 65 years. This study proposes a literature-based drug repurposing strategy in Parkinson's disease.

The literature-based drug repurposing strategy proposed herein combined natural language processing, network science and machine learning methods for analyzing unstructured text data and producing actional knowledge for drug repurposing. The approach comprised multiple computational components, including the extraction of biomedical entities and their relationships, knowledge graph construction, knowledge representation learning and machine learning-based prediction.

The proposed strategy was used to mine information pertaining to the mechanisms of disease treatment from known treatment relationships and predict drugs for repurposing against Parkinson's disease. The F1 score of the best-performing method was 0.97, indicating the effectiveness of the proposed approach. The study also presents experimental results obtained by combining the different components of the strategy.

The drug repurposing strategy proposed herein for Parkinson's disease is distinct from those existing in the literature in that the drug repurposing pipeline includes components of natural language processing, knowledge representation and machine learning for analyzing the scientific literature. The results of the study provide important and valuable information to researchers studying different aspects of Parkinson's disease.

The expansion of marketing data is encouraging the growing use of deep learning (DL) in marketing. I summarize the intuition behind deep learning and explain the mechanisms of six popular algorithms: three discriminative (convolutional neural network (CNN), recurrent neural network (RNN), and Transformer), two generative (variational autoencoder (VAE) and generative adversarial networks (GAN)), and one RL (DQN). I discuss what marketing problems DL is useful for and what fueled its growth in recent years. I emphasize the power and flexibility of DL for modeling unstructured data when formal theories and knowledge are absent. I also describe future research directions.