Dynamic Knowledge Research Articles

SceneKG: A Geo‐Scene Based Spatiotemporal Knowledge Representation Framework Considering Geo‐Processes

ABSTRACTCurrent geographic knowledge graph research mainly focuses on static knowledge coupling through analysis of web text and maps, but its knowledge expression model fails to incorporate knowledge mechanisms such as process evolution and interactions. The geo‐scene is a synthesis of various geographic elements which prescribes spatial, temporal, semantic, attribute, and interrelationships. It has demonstrated its utility in facilitating geocomputation, spatiotemporal simulation and reasoning related to geo‐process. Therefore, developing a geo‐scene‐based knowledge graph is significant to improve the ability of spatiotemporal knowledge representation and service. In this study, we first analyze the basic characteristics of geo‐scenes and try to emphasize the evolution of geo‐scenes through redefining the ontologies ‘event’, ‘process’ and ‘state’. To describe the dynamic nested structure of geo‐scenes, we propose a geo‐scene knowledge representation framework, and illustrate the representation of a geo‐scene and its inner relationships. Moreover, to keep the integrity, interaction and dynamic of geo‐scenes, we develop a event‐process centered geo‐scene knowledge organization method. Furthermore, we have tested the utility of our proposed representation by the Beijing 731 heavy rainstorm. The case study represents heavy rainstorm to demonstrate the complicated structure of the disaster scene, and the support for precipitation process simulation, early warning information organization, and disaster situation information query. The proposed representation provides an idea for the construction of geo‐dynamic knowledge graphs, encodes both static and dynamic geographic knowledge, makes explicit the evolution and interaction knowledge in geo‐scenes, and expands the application service of geo‐scene knowledge graph in the era of GeoAI.

Dynamic multi teacher knowledge distillation for semantic parsing in KBQA

Knowledge base question answering (KBQA) is an important task of extracting answers from a knowledge base by analyzing natural language questions. Semantic parsing methods convert natural language questions into executable logical forms to obtain answers on the knowledge base. Conventional approaches often prioritize singular logical forms, overlooking the distinct strengths inherent in various logical frameworks for problem solving. Recognizing that different logical forms may excel in addressing specific types of questions, our aim is to harness these strengths. By integrating the strengths of different logical forms, we expect to achieve more comprehensive and effective semantic parsing solutions. In our paper, we propose a Dynamic Multi Teacher Knowledge Distillation for Semantic Parsing (DMTKD-SP). DMTKD-SP leverages a collection of teacher models, each mastering a unique logical form, to collaboratively guide a student model so that knowledge from different logical forms can be transferred into the student model. To achieve this, we employ a confidence-based weight assignment module to dynamically assign weights for each teacher model. Furthermore, we introduce a self-distillation mechanism to mitigate the confusion caused by simultaneous learning from multiple teachers. We evaluate DMTKD-SP across variants of the KQA Pro dataset, demonstrating an accuracy improvement of 0.35% on five types of questions, with a notable 0.75% improvement for Count questions.

Crafting innovative paths in non-linear professional learning for bilingual education: the role of connectivism in the age of AI

ABSTRACT This study introduces a novel Connectivist model for professional development (PD) aimed at enhancing AI literacy among bilingual/ESL teachers in K-12 education. As AI technologies rapidly reshape educational landscapes, effective PD programs are essential to equip educators with the skills to integrate AI responsibly and equitably. Grounded in Connectivism, which emphasizes non-linear learning and dynamic knowledge networks, this research explores how PD can transform educators’ perceptions of AI from apprehension to empowerment. The study involved a two-phase PD program with 30 elementary teachers and administrators, incorporating in-person AI training sessions, asynchronous tutorials, and collaborative activities. Data were collected through a phenomenological approach to capture educators’ lived experiences, with triangulation methods using surveys, interviews, observations, and Q&A sessions. The analysis revealed significant improvements in participants’ attitudes toward AI and their openness to integrating it into their teaching practices. The Connectivist model developed in this study fosters ongoing collaborative learning networks and equips educators to leverage AI in addressing resource gaps in bilingual education. This model offers a transformative, sustainable, and adaptive framework for integrating AI into K-12 education, providing a forward-looking approach to professional development in the digital age.

Using reflection and dialog feedback to promote the development of situated and dynamic pedagogical content knowledge

This study explores the reflection and dialog feedback of preservice teachers (PTs) in a one-semester chemistry teaching design and practice course with the aim of improving their situated and dynamic pedagogical content knowledge (PCK). Using a collective case study method, reflections and dialogs from 18 PTs across six teaching cycles were analyzed. The research results indicate that teaching reflection focuses on five themes: the diversity of student situations, the functionality of teaching contexts, the consistency of teaching content, the construction of teacher‒student dialog, and the progression of subject practice activities. The reflection of the PTs strengthened meaningful teaching content and corresponding specific skills, and effective dialog feedback provided external evaluation standards. The reflection prompted teachers to evaluate from multiple perspectives and levels and provided important insights to promote the development of situated and dynamic PCK. This study concludes with a discussion of the practical implications and potential for future research to develop a situated and dynamic PCK for PTs.

Learning‐based tracking control of AUV: Mixed policy improvement and game‐based disturbance rejection

AbstractA mixed adaptive dynamic programming (ADP) scheme based on zero‐sum game theory is developed to address optimal control problems of autonomous underwater vehicle (AUV) systems subject to disturbances and safe constraints. By combining prior dynamic knowledge and actual sampled data, the proposed approach effectively mitigates the defect caused by the inaccurate dynamic model and significantly improves the training speed of the ADP algorithm. Initially, the dataset is enriched with sufficient reference data collected based on a nominal model without considering modelling bias. Also, the control object interacts with the real environment and continuously gathers adequate sampled data in the dataset. To comprehensively leverage the advantages of model‐based and model‐free methods during training, an adaptive tuning factor is introduced based on the dataset that possesses model‐referenced information and conforms to the distribution of the real‐world environment, which balances the influence of model‐based control law and data‐driven policy gradient on the direction of policy improvement. As a result, the proposed approach accelerates the learning speed compared to data‐driven methods, concurrently also enhancing the tracking performance in comparison to model‐based control methods. Moreover, the optimal control problem under disturbances is formulated as a zero‐sum game, and the actor‐critic‐disturbance framework is introduced to approximate the optimal control input, cost function, and disturbance policy, respectively. Furthermore, the convergence property of the proposed algorithm based on the value iteration method is analysed. Finally, an example of AUV path following based on the improved line‐of‐sight guidance is presented to demonstrate the effectiveness of the proposed method.

Instance-Level Scaling and Dynamic Margin-Alignment Knowledge Distillation for Remote Sensing Image Scene Classification

Instance-Level Scaling and Dynamic Margin-Alignment Knowledge Distillation for Remote Sensing Image Scene Classification

MeSHelper: Predicting the evolution of Medical Subject Headings based on knowledge graph dynamics

The Medical Subject Headings (MeSH) thesaurus is a controlled vocabulary widely used in biomedical knowledge systems. We propose a novel framework, termed MeSHelper, that employs a dynamic knowledge graph to predict whether the MeSH main headings (MHs) will evolve while also predicting their corresponding revision type. We parsed the whole PubMed database and all MeSH releases to construct a dynamic semantic tree (DST) and a dynamic knowledge network (DKN) to characterise the evolutionary patterns of MHs and create prediction models. Our results show that DST-related features play a major role in predicting whether the MHs will be revised. Our prediction performance achieved an F1 score of 92.07%. Both DST- and DKN-related features play a crucial role in predicting which types of MHs will evolve. The prediction performance achieved a Macro-F1 score of 72.15%, a Micro-F1 score of 84.09% and a Weighted-F1 score of 84.55%. The findings of this work aid both in constructing an automatic update model for domain thesauruses and in detecting evolutionary trends of the domain knowledge system.

The Cycles agroecosystem model: Fundamentals, testing, and applications

Models of the soil–plant-atmosphere continuum are repositories of knowledge and gears in analytical and decision support tools applied to agroecosystems. In this paper, we present the Cycles agroecosystem model theory along with test cases and applications. Cycles combines innovations for simulating soil hydrology and biogeochemistry, including carbon and nitrogen saturation theory, with a modular software architecture. These elements enable simulating monoculture or polyculture crop sequences and associated management practices, containerizing for applications that require high-performance computing, and data assimilation at runtime. A comparison of simulated and measured daily evapotranspiration (ET) obtained with the eddy covariance method for maize (Zea mays L.) and shrub willow (Salix spp.) shows that Cycles represents well meteorological and vegetation controls of ET (root mean square error or RMSE = 0.75 mm d-1). Cycles accurately simulated differences of 150 mm in growing season ET between these two plant communities. Comparisons of modeled versus measured soil water content under soybean (Glycine max [L.] Merr.) in southeastern Pennsylvania for six soil layers at 0.1-m increments show accurate representation of water depletion and recharge (RMSE of 0.027–0.011 m3 m−3). Simulations of growth and nitrogen uptake of wheat (Triticum aestivum L.) in eastern Washington also highlight the model’s skill simulating processes that affect water and nutrient fluxes simultaneously. To highlight Cycles’ suitability for incorporation in high performance computing applications, we present a coupling of Cycles with an autonomous crop sequence builder (Cycles-A) in the Chesapeake Bay watershed. This system automatically identified areas for double cropping and selected the optimum combination of annual crops across the watershed. The Cycles model innovations and agroecosystem framing continue advancing the premise of making models not only dynamic knowledge repositories but useful tools for research and landscape management.

Dynamic collaborative learning with heterogeneous knowledge transfer for long-tailed visual recognition

Solving the long-tailed visual recognition with deep convolutional neural networks is still a challenging task. As a mainstream method, multi-experts models achieve SOTA accurary for tackling this problem, but the uncertainty in network learning and the complexity in fusion inference constrain the performance and practicality of the multi-experts models. To remedy this, we propose a novel dynamic collaborative learning with heterogeneous knowledge transfer model (DCHKT) in this paper, in which experts with different expertise collaborate to make predictions. DCHKT consists of two core components: dynamic adaptive weight adjustment and heterogeneous knowledge transfer learning. First, the dynamic adaptive weight adjustment is designed to shift the focus of model training between the global expert and domain experts via dynamic adaptive weight. By modulating the trade-off between the learning of features and classifier, the dynamic adaptive weight adjustment can enhance the discriminative ability of each expert and alleviate the uncertainty of model learning. Then, heterogeneous knowledge transfer learning, which measures the distribution differences between the fusion logits of multiple experts and the predicted logits of each expert with different specialties, can achieve message passing between experts and enhance the consistency of ensemble prediction in model training and inference to promote their collaborations. Finally, extensive experimental results on public long-tailed datasets: CIFAR-LT, ImageNet-LT, Place-LT and iNaturalist2018, demonstrate the effectiveness and superiority of our DCHKT.

Research on Construction of Knowledge Base for MRI Equipment Maintenance Based on Knowledge Graph

To improve the level of modern hospital management and establish a scientific and effective medical equipment maintenance guarantee system, this paper proposes the construction of a dynamic interactive fault repair knowledge base system for magnetic resonance imaging (MRI) equipment based on knowledge graph technology. This innovative application of intelligent management for large medical equipment involves collecting and organizing actual maintenance data. The paper constructs the medical equipment maintenance knowledge base, with MRI equipment as a case study, and initially implements functions such as fault information retrieval, technical guidance, and training and learning for MRI equipment. These functions aim to provide an objective decision-making basis for the maintenance and management of large medical equipment, enhance the efficiency of fault repair and maintenance, and elevate the construction level of intelligent hospitals.

Dynamic Routing and Knowledge Re-Learning for Data-Free Black-Box Attack.

Deep learning models have emerged as strong and efficient tools that can be applied to a broad spectrum of complex learning problems and many real-world applications. However, more and more works show that deep models are vulnerable to adversarial examples. Compared to vanilla attack settings, this paper advocates a more practical setting of data-free black-box attack, for which the attackers can completely not access the structures and parameters of the target model, as well as the intermediate features and any training data associated with the model. To tackle this task, previous methods generate transferable adversarial examples from a transparent substitute model to the target model. However, we found that these works have the limitations of taking static substitute model structure for different targets, only using hard synthesized examples once, and still relying on data statistics of the target model. This may potentially harm the performance of attacking the target model. To this end, we propose a novel Dynamic Routing and Knowledge Re-Learning framework (DraKe) to effectively learn a dynamic substitute model from the target model. Specifically, given synthesized training samples, a dynamic substitute structure learning strategy is proposed to adaptively generate optimal substitute model structure via a policy network according to different target models and tasks. To facilitate the substitute training, we present a graph-based structure information learning to capture the structural knowledge learned from the target model. For the inherent limitation that online data generation can only be learned once, a dynamic knowledge re-learning strategy is proposed to adjust the weights of optimization objectives and re-learn hard samples. Extensive experiments on four public image classification datasets and one face recognition benchmark are conducted to evaluate the efficacy of our Drake. We can obtain significant improvement compared with state-of-the-art competitors. More importantly, our DraKe consistently achieves attack superiority for different target models (e.g., residual networks, and vision transformers), showing great potential for complex real-world applications.

G4LDB 3.0: a database for discovering and studying G-quadruplex and i-motif ligands.

Non-canonical nucleic acid structures, such as G-quadruplex (G4) and i-Motif (iM), have garnered significant research interest because of their unique structural properties and biological activities. Thousands of small molecules targeting G4/iM structures have been developed for various chemical and biological applications. In response to the growing interest in G4-targeting ligands, we launched the first G4 Ligand Database (G4LDB) in 2013. Here, we introduce G4LDB 3.0 (http://www.g4ldb.com), an upgraded version featuring extensive enhancements in content and functionality. The new version includes over 4800 G4/iM ligands and approximately 51 000 activity entries. Key upgrades include advanced search capabilities, dynamic knowledge graphs, enhanced data visualization, along with a new dynamic analysis function that automatically displays ligand structure clustering results and chemical space distribution. With these updates, G4LDB 3.0 further evolves into a comprehensive resource and valuable research tool. The significant improvements address the increasing demand for efficient data handling and user experience, highlighting the critical role of G4LDB in advancing research on G-quadruplexes and i-motifs.

Can family structure and contact with natural resources influence young people’s knowledge about medicinal plants? An approach in the Northeast of Brazil

BackgroundIndividuals develop crucial survival knowledge in the juvenile phase, including understanding medicinal plants. The family context or contact with resources can influence this dynamic knowledge. By investigating the influence of these factors on young people's understanding of medicinal plants, we aimed to enhance our understanding of the knowledge-building process.MethodsThe study was conducted in three communities in the State of Alagoas, Brazil: Lagoa do Junco, Serrote do Amparo, and Brivaldo de Medeiros. Semi-structured interviews were conducted with young people aged between 11 and 19 to assess their knowledge of medicinal plants. We used a generalized linear mixed model (GLMM) with the number of therapeutic targets and known medicinal species (knowledge proxy) as response variables. As predictors, we included the number of individuals per family unit and the gender distribution within the famimunícpily (family context proxy), as well as dependence on the use (contact proxy). Location (city) was added as a fixed effect to the model. We investigated how knowledge of medicinal plants correlates with the practice of collecting these species.ResultsWe did not identify a relationship between the number of individuals per household, gender distribution within the family, and the frequency of medicinal plant use with knowledge about these species. However, we observed a positive, albeit weak, correlation between knowledge of medicinal species and the number of species collected.ConclusionThese results highlight the importance of investigating how young people acquire knowledge about medicinal plants, emphasizing the complex interactions between humans and nature, and providing a basis for future research.

Automatic deep sparse clustering with a dynamic population-based evolutionary algorithm using reinforcement learning and transfer learning

Clustering data effectively remains a significant challenge in machine learning, particularly when the optimal number of clusters is unknown. Traditional deep clustering methods often struggle with balancing local and global search, leading to premature convergence and inefficiency. To address these issues, we introduce ADSC-DPE-RT (Automatic Deep Sparse Clustering with a Dynamic Population-based Evolutionary Algorithm using Reinforcement Learning and Transfer Learning), a novel deep clustering approach. ADSC-DPE-RT builds on Multi-Trial Vector-based Differential Evolution (MTDE), an algorithm that integrates sparse auto-encoding and manifold learning to enable automatic clustering without prior knowledge of cluster count. However, MTDE's fixed population size can lead to either prolonged computation or premature convergence. Our approach introduces a dynamic population generation technique guided by Reinforcement Learning (RL) and Markov Decision Process (MDP) principles. This allows for flexible adjustment of population size, preventing premature convergence and reducing computation time. Additionally, we incorporate Generative Adversarial Networks (GANs) to facilitate dynamic knowledge transfer between MTDE strategies, enhancing diversity and accelerating convergence towards the global optimum. This is the first work to address the dynamic population issue in deep clustering through RL, combined with Transfer Learning to optimize evolutionary algorithms. Our results demonstrate significant improvements in clustering performance, positioning ADSC-DPE-RT as a competitive alternative to state-of-the-art deep clustering methods.

The relationship between entrepreneurial personality patterns linked to risk, innovation and gender across industrial sectors

This study examines the personality patterns of solo founders in both high-tech and non-high-tech sectors during the first seven years of their entrepreneurial journey to emphasize the patterns’ implications during policymaking, investment decisions, and self-assessments. IAB/ZEW startup panel microdata for the sector classification of 4470 solo entrepreneurs in Germany were analyzed to identify Big Five trait patterns influenced by risk propensities, innovation inclination, and gender. The entrepreneurial profiles indicate positive openness, emotional resilience, and sector-specific clusters. Conscientiousness suggests flexibility, and while variations in extraversion and agreeableness exist, negative neuroticism was predominantly found, except for gender-related differences and multidimensional service innovators. Big Five traits provide information about important foundational profile patterns to describe unique solo entrepreneur types influenced by risk, innovation, and gender. Originality and value: Risk propensity characterizes ‘Adaptive Services,’ ‘Dynamic Knowledge Innovators,’ and ‘Strategic Risk Navigators.’ Additionally, ‘Multidimensional Service Innovators’ and ‘Focused Tech Innovators’ signify innovation understanding. The Big Five profiles show openness and emotional stability across sectors, providing crucial insights for effective entrepreneurial support and investment strategies.

Urban Vulnerability Assessment of Sea Level Rise in Singapore through the World Avatar

This paper explores the application of The World Avatar (TWA) dynamic knowledge graph to connect isolated data and assess the impact of rising sea levels in Singapore. Current sea level rise vulnerability assessment tools are often regional, narrow in scope (e.g., economic or cultural aspects only), and are inadequate in representing complex non-geospatial data consistently. We apply TWA to conduct a multi-perspective impact assessment of sea level rise in Singapore, evaluating vulnerable buildings, road networks, land plots, cultural sites, and populations. We introduce OntoSeaLevel, an ontology to describe sea level rise scenarios, and its impact on broader elements defined in other ontologies such as buildings (OntoBuiltEnv ontology), road networks (OpenStreetMap ontology), and land plots (Ontoplot and Ontozoning ontology). We deploy computational agents to synthesise data from government, industry, and other publicly accessible sources, enriching buildings with metadata such as property usage, estimated construction cost, number of floors, and gross floor area. An agent is applied to identify and instantiate the impacted sites using OntoSeaLevel. These sites include vulnerable buildings, land plots, cultural sites, and populations at risk. We showcase these sea level rise vulnerable elements in a unified visualisation, demonstrating TWA’s potential as a planning tool against sea level rise through vulnerability assessment, resource allocation, and integrated spatial planning.

Exploring digital transformation strategy to achieve SMEs resilience and antifragility: a systematic literature review

Nowadays, the business environment has become more dynamic, making survival issues more challenging for small and medium enterprises (SMEs). Academic literature proposes digital transformation as a facilitator for SMEs to generate resilience and antifragility to overcome this challenge. However, SMEs need appropriate strategies to be successful in their digital transformation journey. This study aims to construct a digital transformation strategy framework for SMEs to generate resilience and antifragility. We use systematic literature review (SLR) to capture critical knowledge from the published literature. The primary articles were analyzed using thematic analysis with Wolcott’s (1994) procedure to construct the framework based on the primary studies. We found the critical values SMEs need to achieve successful digital transformation, including dynamic capabilities, digital capability, digital inclusion, leadership orientation, learning and knowledge management, and collaboration. However, SMEs need a deeper level of learning, higher digital capability, flexibility, and agility to be antifragile. Furthermore, we found that dynamic capabilities are the leading theory used to describe and investigate how a firm generates successful transformation. Finally, this study proposes a conceptual framework for digital transformation strategies to generate SME resilience and antifragile by connecting theory and practical concepts. It also suggests future research agendas.

MLC-DKT: A multi-layer context-aware deep knowledge tracing model

As a fundamental research task for intelligent education, knowledge tracing (KT) aims to trace learners’ dynamic knowledge states and predict their future performance based on their historical response data. However, learners’ cognitive processes cannot be studied in isolation from the context; nevertheless, existing knowledge tracing models mainly use knowledge topology as auxiliary information but do not deeply explore the contextual information of exercises. Thus, an in-depth exploration of multi-layer contextual information in learning scenarios is required, focusing on the cognitive differences among different learners. Furthermore, knowledge tracing models exhibit limited interpretability compared with cognition diagnosis methods integrating educational priors. To address these issues, we propose a multi-layer context-aware deep knowledge tracing (MLC-DKT) model. Specifically, we first present the multi-layer contextual representation method involving knowledge concepts and exercises. In the knowledge state tracing module, we incorporate the learners’ knowledge priors into the attention mechanism to capture the evolution of the learners’ knowledge. Then, we develop a calculation method for the similarity effect among the contextual information, which helps estimate the learners’ knowledge states. In the learning performance prediction module, we introduce the guessing and slipping factors from the cognition diagnosis model to optimize the MLC-DKT model, further improving the performance of the KT model. Finally, we conduct extensive experiments on real-world datasets, demonstrating that the MLC-DKT model provides improved predictions of learner performance. Moreover, the MLC-DKT model realizes contextual awareness from knowledge concepts and exercises, making the predictions more interpretable.

Leveraging local knowledge for landslide disaster risk reduction in an urban informal settlement in Manado, Indonesia

Disaster preparedness and response, including for landslides, increasingly involves local knowledge. Incorporating contextual, dynamic and experience-based local knowledge leads to greater awareness of the interconnectedness of geological, natural, and social processes. Still, technical literature on urban landslide risk is mainly based on geological and natural dynamics and, to a lesser extent, physical infrastructure. Moreover, although local knowledge is recognized as important in principle for all aspects of disaster risk management, it is not frequently taken as a starting point for identifying contextually specific landslide risk factors in practice. To gain insights into local knowledge on factors contributing to landslide risk, we conducted a qualitative case study within a landslide-prone informal settlement in Manado, Indonesia. The study comprised qualitative household interviews, transect walks, and ethnographic observation. Our findings indicate that anthropogenic factors specific to informal settlements, particularly inadequate solid waste management, may serve as an overlooked risk factor for urban landslides. This is especially problematic in low-income or informal neighborhoods, where detrimental effects of poor solid waste management frequently intersect with heightened vulnerability to various hazards. We therefore advocate systematically incorporating anthropogenic factors into studies of contributing factors to urban landslides. Additionally, we encourage further research into the interactions between inadequate solid waste management and landslide risk. Finally, we call for leveraging local knowledge to enhance policy, planning, and research efforts, with the ultimate goal of fostering safer urban environments for all.

A Demonstration of TENDS: Time Series Management System Based on Model Selection

The growth in sensor technologies, IoT devices, and information systems has opened up new opportunities for managing time series data across various domains. Despite significant progress, existing time series management systems face two crucial limitations: insufficient functionality and inadequate adaptability. This highlights the need for more comprehensive systems that not only improve data quality and analysis but also effectively manage the variety and volume of time series data. This paper presents TENDS, a time series management system based on model selection. TENDS uniquely combines advanced functionalities for imputation, prediction, and anomaly detection, offering a comprehensive analysis of time series data. It features (i) an effective model selection mechanism to adapt to various data types and to improve efficiency; (ii) fourteen state-of-the-art prediction methods and three state-of-the-art imputation methods; and (iii) a dynamic expert knowledge base for anomaly detection, evolving continuously with new data to ensure accuracy. TENDS boasts a comprehensive suite of visualization tools. With its configurable offline and online interfaces, TENDS (i) provides extensive flexibility in model selection and parameter adjustment, (ii) facilitates easy visualization of training results, and (iii) supports real-time documentation and statistical analysis of time series.