Joint Knowledge Research Articles

Communication-Efficient and Privacy-Preserving Federated Learning via Joint Knowledge Distillation and Differential Privacy in Bandwidth-Constrained Networks

Communication-Efficient and Privacy-Preserving Federated Learning via Joint Knowledge Distillation and Differential Privacy in Bandwidth-Constrained Networks

Construction of Legal Knowledge Graph Based on Knowledge-Enhanced Large Language Models

Legal knowledge involves multidimensional heterogeneous knowledge such as legal provisions, judicial interpretations, judicial cases, and defenses, which requires extremely high relevance and accuracy of knowledge. Meanwhile, the construction of a legal knowledge reasoning system also faces challenges in obtaining, processing, and sharing multisource heterogeneous knowledge. The knowledge graph technology, which is a knowledge organization form with triples as the basic unit, is able to efficiently transform multisource heterogeneous information into a knowledge representation form close to human cognition. Taking the automated construction of the Chinese legal knowledge graph (CLKG) as a case scenario, this paper presents a joint knowledge enhancement model (JKEM), where prior knowledge is embedded into a large language model (LLM), and the LLM is fine-tuned through the prefix of the prior knowledge data. Under the condition of freezing most parameters of the LLM, this fine-tuning scheme adds continuous deep prompts as prefix tokens to the input sequences of different layers, which can significantly improve the accuracy of knowledge extraction. The results show that the knowledge extraction accuracy of the JKEM in this paper reaches 90.92%. Based on the superior performance of this model, the CLKG is further constructed, which contains 3480 knowledge triples composed of 9 entities and 2 relationships, providing strong support for an in-depth understanding of the complex relationships in the legal field.

Intelligent maritime question-answering and recommendation system based on maritime vessel activity knowledge graph

Traditional maritime traffic management typically relies on positioning data for data mining without incorporating other multi-source data to analyze the maritime vessel activity, which cannot conduct comprehensive maritime knowledge mining. Thus, this study integrates multi-source data, such as trajectory, maritime accident text, and geographic data, to create a maritime vessel activity knowledge graph. On this basis, a question-answering model is developed based on a bidirectional question-answering attention graph neural network, and a personalized recommendation model is developed based on an attention-enhanced joint knowledge propagation and a user preference graph neural network. The former assists users in extracting valuable information from the maritime vessel activity knowledge graph, while the latter predicts the users' potential interests and automatically recommends vessel entities based on their historical query information. Experimental results show that the proposed question-answering model improved the F1-score by 2.31%–10.09% compared to state-of-the-art baseline models on the MVA question-answering dataset. Similarly, the proposed personalized recommendation model improved the click-through rate prediction accuracy by 2.46%–7.05% compared to state-of-the-art baseline models on the MVA personalized recommendation dataset.

Construction of Cultural Heritage Knowledge Graph Based on Graph Attention Neural Network

To address the challenges posed by the vast and complex knowledge information in cultural heritage design, such as low knowledge retrieval efficiency and limited visualization, this study proposes a method for knowledge extraction and knowledge graph construction based on graph attention neural networks (GAT). Using Tang Dynasty gold and silver artifacts as samples, we establish a joint knowledge extraction model based on GAT. The model employs the BERT pretraining model to encode collected textual knowledge data, conducts sentence dependency analysis, and utilizes GAT to allocate weights among entities, thereby enhancing the identification of target entities and their relationships. Comparative experiments on public datasets demonstrate that this model significantly outperforms baseline models in extraction effectiveness. Finally, the proposed method is applied to the construction of a knowledge graph for Tang Dynasty gold and silver artifacts. Taking the Gilded Musician Pattern Silver Cup as an example, this method provides designers with a visualized and interconnected knowledge collection structure.

Optimizing Collaboration Between OTs & Teachers in Inclusive Education: Key Considerations for a Joint Knowledge Translation Intervention

Abstract Date Presented 03/23/24 This presentation will highlight key considerations related to collaboration between OTs and teachers in inclusive education settings. These findings inform the development of a joint knowledge transfer training to optimize collaborative practices. Primary Author and Speaker: Lina Ianni Contributing Authors: Dana Anaby, Chantal Camden

A joint knowledge representation learning of sentence vectors weighting and primary neighbor constraints

A joint knowledge representation learning of sentence vectors weighting and primary neighbor constraints

Deformation ability of steel inner sleeve T-joint in modular gymnasia based on 3D-DIC method

Substantial loads the roofs bear exert stringent performance standards on connecting joints in movable modular gymnasiums. Previously, we introduced a novel joint design, utilizing internal sleeves and bolts to integrate the upper and lower columns of modular units. To explore the actual deformation behavior and force distribution of this joint, this study employed a three-dimensional digital image correlation (3D-DIC) measurement method to visualize its full-field displacement and strain distribution. The reliability of the 3D-DIC results was confirmed through comparison with traditional measurement methods. Building upon this solid foundation, this study proceeds to assess the static performance of this innovative joint in two distinct directions, successfully revealing micro-deformation and local buckling phenomena undetectable to the naked eye. Notably, the quality of the welds connecting the beams and columns plays a pivotal role in ensuring the overall structural integrity of the joint. Under lateral loading conditions, the destructive behavior of the joints is primarily governed by material strength. Moment-rotation curves were plotted to delve further into the mechanical characteristics of the joint, revealing that the inner-sleeve T-joint exhibits the attributes of a rigid joint. Moreover, the primary damage modes observed in the specimens were fractures at the beam root or buckling of the column, with the joint area remaining intact. Consequently, this joint qualifies as a full-strength connection. These findings will deepen knowledge and understanding of this novel joint for designers.

Joint Knowledge Graph and Large Language Model for Fault Diagnosis and Its Application in Aviation Assembly

Joint Knowledge Graph and Large Language Model for Fault Diagnosis and Its Application in Aviation Assembly

Discriminative Joint Knowledge Transfer With Online Updating Mechanism for EEG-Based Emotion Recognition

Discriminative Joint Knowledge Transfer With Online Updating Mechanism for EEG-Based Emotion Recognition

Deep Semantic Graph Transformer for Multi-View 3D Human Pose Estimation

Most Graph Convolutional Networks based 3D human pose estimation (HPE) methods were involved in single-view 3D HPE and utilized certain spatial graphs, existing key problems such as depth ambiguity, insufficient feature representation, or limited receptive fields. To address these issues, we propose a multi-view 3D HPE framework based on deep semantic graph transformer, which adaptively learns and fuses multi-view significant semantic features of human nodes to improve 3D HPE performance. First, we propose a deep semantic graph transformer encoder to enrich spatial feature information. It deeply mines the position, spatial structure, and skeletal edge knowledge of joints and dynamically learns their correlations. Then, we build a progressive multi-view spatial-temporal feature fusion framework to mitigate joint depth uncertainty. To enhance the pose spatial representation, deep spatial semantic feature are interacted and fused across different viewpoints during monocular feature extraction. Furthermore, long-time relevant temporal dependencies are modeled and spatial-temporal information from all viewpoints is fused to intermediately supervise the depth. Extensive experiments on three 3D HPE benchmarks show that our method achieves state-of-the-art results. It can effectively enhance pose features, mitigate depth ambiguity in single-view 3D HPE, and improve 3D HPE performance without providing camera parameters. Codes and models are available at https://github.com/z0911k/SGraFormer.

Small Group Conversations in a POGIL-Based Class: How English Learners Engage in a Joint Knowledge Construction Process to Reach a Shared Understanding.

Collaboration is an aspect of engagement that focuses on learning through group work and having discussions with other learners. Active learning approaches are a way to foster collaborative engagement because they provide more opportunities for interaction among learners. Process Oriented Guided Inquiry Learning (POGIL), a socially mediated active learning approach, uses verbalizing and discussing ideas with peers in small groups to reach a shared understanding. Due to the growing number of immigrants in the United States, the number of English Learners (ELs) in American classrooms has been increasing rapidly. ELs encounter challenges such as unfamiliarity with American science class norms and expectations, feelings of not being valued and socially accepted, and instructors' lower expectations of them. These challenges can impact this group of students' learning and achievements. Previous studies have discussed that learning chemistry can be more challenging for EL students due to the critical role of language in learning. We argue that ELs use discourse moves differently compared to non-ELs in a POGIL-based class in terms of engaging in conversations that can lead to a shared understanding at the group level through a joint knowledge construction process. Our findings indicated that, in our sample, ELs were less likely to engage in discursive moves than non-ELs. This difference may result in missed opportunities for a shared understanding and joint knowledge construction. In addition to differences between EL and non-EL students in our samples, we also found differences between EL students who attended K-12 schools in the United States compared to international EL students. Implications for future studies of these possibly distinct EL populations are considered.

Building Common Ground: How Facilitators Bridge Between Diverging Groups in Multi-Stakeholder Dialogue

The effectiveness of multi-stakeholder initiatives (MSIs) in tackling grand social and environmental challenges depends on productive dialogue among diverse parties. Facilitating such dialogue in turn entails building common ground in form of joint knowledge, beliefs, and suppositions. To explore how such common ground can be built, we study the role of different facilitators and their strategies for bridging the perspectives of competing stakeholder groups in two contrasting MSIs. The German Partnership for Sustainable Textiles was launched in an initially hostile communicative environment, whereas the Fossil Free Sweden Initiative proceeded in a fertile communicative environment. We trace how the facilitators in these initiatives achieved common ground through three bridging strategies—communicative integration, temporal calibration, and process alignment—adapted to the communicative environments of these MSIs. In hostile communicative environments, facilitators achieve common ground by steering diverging stakeholder groups towards ‘reconciling’ their different language registers, knowledge bases, and meaning systems to ‘meet in the middle’ on points of agreement and shared interests. In fertile communicative environments characterised by greater mutual trust, facilitators can steer interactants to ‘strategically appropriate’ to the language, knowledge, and meaning system of a particular stakeholder group to win this group’s support. Our analysis contributes to a better understanding of how productive multi-stakeholder dialogue can be facilitated.

Knowledge Distillation Using Hierarchical Self-Supervision Augmented Distribution.

Knowledge distillation (KD) is an effective framework that aims to transfer meaningful information from a large teacher to a smaller student. Generally, KD often involves how to define and transfer knowledge. Previous KD methods often focus on mining various forms of knowledge, for example, feature maps and refined information. However, the knowledge is derived from the primary supervised task, and thus, is highly task-specific. Motivated by the recent success of self-supervised representation learning, we propose an auxiliary self-supervision augmented task to guide networks to learn more meaningful features. Therefore, we can derive soft self-supervision augmented distributions as richer dark knowledge from this task for KD. Unlike previous knowledge, this distribution encodes joint knowledge from supervised and self-supervised feature learning. Beyond knowledge exploration, we propose to append several auxiliary branches at various hidden layers, to fully take advantage of hierarchical feature maps. Each auxiliary branch is guided to learn self-supervision augmented tasks and distill this distribution from teacher to student. Overall, we call our KD method a hierarchical self-supervision augmented KD (HSSAKD). Experiments on standard image classification show that both offline and online HSSAKD achieves state-of-the-art performance in the field of KD. Further transfer experiments on object detection further verify that HSSAKD can guide the network to learn better features. The code is available at https://github.com/winycg/HSAKD.

Supply chain integration practices and its impact on financial and operational performance of the Tunisian industrial sector

Supply chain integration is a key factor in improving business performance. Despite this growing interest, little research has highlighted this issue that will determine the survival of many companies. The purpose of this study was to examine the relationship between the supply chain integration practices (SCMP) and the financial and operational performance of Tunisian manufacturing companies. To achieve this, a questionnaire was used as a research tool for data collection. Multiple regression analysis using SPSS26 software was used to answer study questions and examine study hypotheses. According to research, SCMP includes customer relationship management (CRM), supplier relationship management (SRM), buyer-supplier collaboration (BSC), joint knowledge creation (JCK), level of information sharing (IS), goal congruence (GC) , risk management (RM), internet usage (IU), quality management (QM) , except information quality (IQ), collaborative planning, forecasting and replenishment practices (CPFR), and just-in-time (JIT), are all significantly related to the financial and operational performance of the companies surveyed. Some recommendations are proposed to help managers better manage knowledge throughout the supply chain.

Theoretical Advancement of Xi Jinping’s Socialist Thought with Chinese Characteristics in the New Era under the Construction of Knowledge Map

Abstract This paper discusses the theory and practice of constructing knowledge graph under the guidance of Xi Jinping’s socialist thought with Chinese characteristics in the new era. The study first adopts the Transformer technique based on the BERT model for effectively extracting entity relationships. Then, the article proposes an entity alignment method based on Neural Tensor Network, which enhances the semantic and structural information utilization of entity embedding vectors through optimized NTN and joint knowledge representation learning. In addition, the study employs a knowledge graph link prediction method that combines semantics and paths to fill in missing relations, and applies probabilistic soft logic to solve the inference problem of non-deterministic knowledge. Ultimately, the effectiveness of the constructed knowledge graph is demonstrated through empirical analysis of Xi Jinping’s theoretical dataset of socialist thought with Chinese characteristics in the new era. The results show that the method is more effective in dealing with polysemous words and participle errors than traditional methods, improving the data structuring level. For example, the accuracy of entity alignment is significantly better than the baseline algorithm on different datasets, e.g., on the DBP-WE dataset, its Hits@1, Hits@10 and MRR values are higher than those of the baseline by 4.01-29.49, 1.99-31.05, and 0.036-0.289, respectively. The method of this paper is remarkably effective in improving the performance of the entity alignment task, which is helpful in understanding and applying the Xi Jinping’s thought of socialism with Chinese characteristics in the new era provides new theoretical perspectives and practical methods.

Epistemic Responsibility and Community Engagement in Sindiwe Magona’s Beauty’s Gift

HIV/AIDS has pervasively affected the health and well-being of South African women, as evidenced by their exacerbating mortality rates over the decades. Sindiwe Magona’s Beauty’s Gift (2008) is a critical intervention in this regard because, in focusing on the death of a young black woman owing to her infection with HIV/AIDS, it critiques the overarching and ingrained patriarchal ideologies that are hindering the treatment and prevention of this disease. The article seeks to examine Magona’s text as uncovering the lack of appropriate knowledge about HIV/AIDS, and the silencing of South African women in the articulation or dissemination of this disease-related knowledge. The article argues that Magona’s text emphasizes the delinking of taboos and highlights the necessity of the South African community’s engagement at micro and macro levels regarding HIV/AIDS. This engagement reflects an epistemic responsibility that is crucial to reducing the dreadful impact of this illness from both cultural and structural perspectives.

Leveraging Tacit Knowledge for Teacher Performance in Vocational Schools After-Break

Purpose – The primary objective of this study is to examine the utility of tacit knowledge in enhancing the performance of vocational school teachers who have experienced a career hiatus. Additionally, it seeks to investigate strategies for vocational schools to provide improved support to their teachers in leveraging tacit knowledge to enhance performance, particularly through the promotion of work engagement. Furthermore, the study aims to present a framework for vocational school managers, facilitating the explicit articulation of teachers' tacit knowledge, thereby contributing to the creation and dissemination of new knowledge through effective knowledge management. Design/methodology/approach – In order to gather data for this research, an explanatory survey methodology is employed, with the study's focus directed towards vocational education teachers in the region of West Java. A meticulously chosen sample size of 360 teachers is selected, guided by specific experience criteria and employment within the foundation overseeing the Private vocational school. Structural equation modeling (SEM) serves as the chosen method for hypothesis testing in this study. Findings – This study finds that teacher work engagement serves as a crucial mediator in the relationship between tacit knowledge and performance among vocational school teachers. It is observed that increased tacit knowledge may unexpectedly lead to reduced teacher performance, emphasizing the complexity of this relationship. Tacit knowledge is identified as a valuable resource for addressing workplace demands within vocational education. Moreover, it exerts a significant influence on the motivation process of teachers, subsequently impacting their engagement and overall performance. Interestingly, engagement plays a mitigating role, counteracting the negative effects of suboptimal tacit knowledge utilization by teachers, allowing them to achieve better performance, although not always reaching their full potential based on their rich experiences. Consequently, work engagement emerges as a key mediator, facilitating the translation of tacit knowledge into effective teaching outcomes. These findings shed light on the intricate dynamics of tacit knowledge, engagement, and performance in vocational education, providing valuable insights for educators and policymakers. Originality/value – The originality of this research manifests in its concentrated exploration of the application of tacit knowledge to enhance the performance of vocational school teachers, particularly those who have taken a hiatus in their careers. Additionally, the study introduces a fresh perspective to the realm of vocational education by highlighting the potential outcomes of making teachers' tacit knowledge explicit, with a particular emphasis on work engagement and knowledge management. This novel approach contributes valuable insights to the field, paving the way for further exploration and development.

Resiliencia y memoria visual. Una mirada educativa al gran terremoto de Valdivia de 1960.

The photographic exhibition Journey to the Epicenter is an exercise of heritage education developed by the Dirección Museológica UACh within the framework of the 60th anniversary of the Valdivia earthquake. The visual understanding of the seismic phenomenon experienced in 1960 is a contribution to the heritage field, by highlighting the value of resilience in the formation of cultural identities, leaving at the disposal of the territories the articulation of scientific, social and emotional knowledge to face catastrophic situations.

Exploring art-making as research enquiry: Methods, materiality and intentionality

This article discusses the intersections of art-making and research enquiry, focusing on methods, materiality and intentionality. In response to the request to contribute to an archive of creative methodological processes, the author examines a methodology he is developing for co-creating artistic acoustic ecologies with the Great Lakes. It raises questions about the role of practice-based research in generating new forms of enquiry and systems of knowledge while examining the barriers to achieving legitimacy for artistic practices within the research landscape. The article explores ethical, technological and artistic considerations within the context of the author’s auto-ethnographic lens of an artist working in an era that demands verbalization as a primary articulation of knowledge and cultural expression. The author, a practising visual artist, musician and Ph.D. student in the research-creation-based Media and Design Innovation programme at Toronto Metropolitan University, discusses the need for concrete examples of how artistic research methods are effective.

Joint knowledge graph approach for event participant prediction with social media retweeting

Organized event is an important form of human activity. Nowadays, many digital platforms offer organized events on the Internet, allowing users to be organizers or participants. For such platforms, it is beneficial to predict potential event participants. Existing work on this problem tends to borrow recommendation techniques. However, compared to e-commerce items and purchases, events and participation are usually of a much smaller frequency, and the data may be insufficient to learn an accurate prediction model. In this paper, we propose to utilize social media retweeting activity to enhance the learning of event participant prediction models. We create a joint knowledge graph to bridge the social media and the target domain, assuming that event descriptions and tweets are written in the same language. Furthermore, we propose a learning model that utilize retweeting information for the target domain prediction more effectively. We conduct comprehensive experiments in two scenarios with real-world data. In each scenario, we set up training data of different sizes, as well as warm and cold test cases. The evaluation results show that our approach consistently outperforms several baseline models in both warm and cold tests.