Knowledge Model Research Articles

Wave masked autoencoder: An electrocardiogram signal diagnosis model based on wave making strategy

ECG signal labeling is time-consuming and laborious, significantly constraining the performance of deep learning models on ECG diagnosis. Self-supervised learning methods offer a means to mitigate the reliance on annotated ECG data. Nevertheless, the direct transposition of these methods to ECG diagnostics may encounter challenges stemming from deficient domain knowledge and incongruous model architectures. Therefore, a self-supervised learning model that introduces ECG prior knowledge, called wave masked autoencoder (WMAE), is proposed. Firstly, considering that various waves that make up the ECG signal are key features for diagnosing arrhythmias, we design a wave masking strategy as an auxiliary task for self-supervised learning to introduce ECG-related knowledge. Secondly, sliding window slicing is utilized to divide each ECG signal into a series of subsequences, which are mapped into deep features with a convolutional feature mapper to increase the information density of basic semantic units and learn complete waveform representations. In this way, the proposed WMAE can capture robust and transferable ECG signal features from unlabeled data, thereby alleviating the problem of insufficient annotated ECG data and improving the performance of the model on cardiovascular disease diagnosis. Experimental results show that WMAE is significantly better than other benchmark models and has high practical value.

Fusing graph structural information with pre-trained generative model for knowledge graph-to-text generation

Fusing graph structural information with pre-trained generative model for knowledge graph-to-text generation

Evaluating geospatial education provision: a case study of Aotearoa New Zealand

ABSTRACT Geospatial technology is still a growth area and knowledge of these technologies remains essential for a skilled workforce. In higher education, geospatial curriculum design is predominantly determined by educators and is influenced by what educators believe should be taught. Knowledge of national or regional provision can support a richer understanding of expected provision and areas of over or under-provision to better support the needs of varied geospatial roles and identify possible programme specialisations. In this paper, we describe an approach for evaluating Aotearoa New Zealand’s provision of geospatial education in the tertiary education sector. First, we examine and compare current course offerings in New Zealand’s tertiary sector against selected institutions in five countries using the GIS&T Body of Knowledge and Geospatial Technology Competency Model as a framework for comparison. Second, we evaluate the results of a focus-group workshop of geospatial education providers and geospatial industry stakeholders about perceived needs in geospatial education in New Zealand. We find this approach works well to support a better understanding of what is provided for both educators and industry. By including industry geospatial professionals in identifying perceived gaps, educators’ knowledge of industry needs will grow as well as industry’s understanding of current provision.

A Comparison of Brand-Related Outcomes Between Naming Rights and Jersey Sponsorships in Professional Sport

Naming rights and jersey sponsorship are highly visible, offering numerous benefits to both brands and properties but at a particularly high cost. The purpose of this study was to examine individuals’ attitudes and behaviors toward naming rights and jersey sponsors. The study drew on the Persuasion Knowledge (PK) Model as a lens for differentiating between consumers’ responses to each sponsor. A between-group design was employed; two versions of a questionnaire (one for each sponsorship condition) were distributed to a nationwide sample of Canadians, yielding 331 usable responses. Findings suggested participants had better awareness of the naming rights sponsor than the jersey sponsor, but no attitudinal differences were observed. However, attitudinal differences were observed on team identification. Furthermore, it appears that the triggering process for PK may differ in the sport sponsorship setting than what is observed with traditional advertising, likely due to a moderating effect of team identification.

Implementing national performance indicator ontology: A case study of Botswana's development agendas

This paper presents the National Performance Indicator Ontology (NPIOnto), a knowledge model designed for the semantic representation of national performance indicators (NPIs) and related performance variables to enhance the semantic capabilities of open data. The research addresses the lack of semantic representation for NPIs, particularly within Botswana's development agendas, which impedes the effective use, accessibility, and integration of performance data essential for monitoring and achieving sustainable development goals. NPIOnto was developed through five steps: discovering the ontology purpose, studying the development agenda knowledge domain, developing the conceptual knowledge model, formalizing the conceptual model, and evaluating the formal model's effectiveness in enhancing the semantic capabilities of open data resources. The evaluation, based on practical application in Botswana, demonstrated that NPIOnto significantly improves the accessibility, understandability, interoperability, integrability, and intelligibility of open data. Insights from the Botswana case study indicate that while NPIOnto is tailored to meet Botswana's development agenda needs, it can be adapted for use in other developing countries.

Reconfigurable flexible assembly model and implementation for cross-category products

As the production orders are becoming multi-category and small-batch in the era of product personalization, these require frequent reconfiguration of reconfigurable flexible assembly system for cross-category products (RFAS-CCP). However, there is no suitable theoretical assembly model and systematic implementation framework. We first propose a five-element assembly model (FAM) for RFAS-CCP, i.e. product, process, resource, knowledge, and decision. The product, process, and resource element describe the objects, steps to be assembled, and the tools, fixtures, and other equipment used for assembly, respectively. The knowledge element is a form representation of various heterogeneous data, such as a knowledge graph. The decision element includes various assembly methods to achieve assembly automation, flexibility, and intelligence. Then, in order to standardize and easy the frequent reconfiguration process, we reorganize various decision methods into a three-phase systematic implementation framework according to which stage they are used: design, configuration, and operation phases. The design phase methods primarily design various assembly modules for a product family, forming an assembly resource library. The configuration phase methods primarily configure suitable assembly lines for a specific product in the product family. The operation phase methods monitor the status of the assembly line and ensures its stable operation through health management. Finally, the effectiveness and practicality of the proposed five-element assembly model and three-phase systematic implementation framework are experimented with a pressure reducing valve product.

Unlocking water sustainability: The role of knowledge, attitudes, and practices among women

Following the knowledge, attitude, and practices model (KAP) this research investigates the influence of knowledge of water sustainability (KWS) on the three attitudes of water sustainability (reduce, reuse, recycle) on sustainable water practices (SWP) among women in Jordan. Likewise, this study examines both the direct and water stewardship (WS) on SWP, and the moderation effect of WS on the relationship between three attitudes of water sustainability: reduce (RED), reuse (REU), and recycle (REC) and SWP. Using Structural Equation Modelling, the results reveal that KWS plays a pivotal role in driving positive influence across the three attitudes of water sustainability. Likewise, all three attitudes of water sustainability have positive influence on driving SWP indicating significant mediation effect in the relationship between KWS and SWP. Importantly, the study highlights a quasi-moderation effect of WS as the results indicate both a significant direct effect of WS on SWP and a significant moderation effect of WS on the relationship between the three sustainable water attitudes and SWP. These findings offer valuable insights for policymakers to design context-specific educational programs and cross-sectoral collaborations, ultimately fostering sustainable water practices and contributing to the global effort to address water scarcity and environmental degradation.

Formation of a modern historiographical model of political and legal knowledge

Formation of a modern historiographical model of political and legal knowledge

Free Text Analysis: Identification of adverse drug events in clinical notes

ObjectiveTo explore the prevalence of adverse drug events (ADE) leading to drug changes using Artificial Intelligence (AI) based free text analysis of primary care encounter notes . ApproachWe used electronic medical record l encounter notes linked to population-based drug dispensation records. Dispensation records were used to identify drug switching to another agent in the same therapeutic class, which could suggest an undesirable side-effect requiring discontinuation of the drug. We annotated clinical notes by identifying ADEs after retrieving the last encounter note written for that patient on or before the identified date. We ran different AI models, including BERT and Large Language Models (LLMs), to identify ADEs. ResultsWe annotated 1085 notes with 362 ADEs. These were used to train BERT models (precision or PPV: 0.584 ) and prompt-based LLM (0.563) but with improved sensitivity (0.725) and specificity (0.739) . ConclusionsIdentifying ADEs requires deep medical knowledge and simple Natural Language Processing models trying to identify ADEs only on the basis of language of encounter notes was not successful based on the BERT model performance. Prompt based models have deep knowledge but were trained with the objective of producing grammatically correct, coherent sentences, not with the medical knowledge required. LLMs also require significant computational power. ImplicationsWhile AI does have significant potential to advance our capacity for complex analytic tasks the current LLMs are not yet able to adequately identify ADEs in a corpus of electronic clinical notes.

The vision of the circular factory for the perpetual innovative product

Abstract The growing scarcity of global resources demands a transition from linear to circular production patterns. This article presents a novel vision for integrating linear and circular production processes within a flexible and autonomous production system to achieve perpetual product use. The approach aims to preserve the added value of products and to integrate the design of product generations and production systems. Within the circular factory, the following core aspects must be examined: predicting functions of products, managing uncertainty in used products and process sequences, learning human action for complex tasks, implementing changeable, autonomous production systems, and enabling knowledge modeling for the circular factory across domains. Aspired results are a design for circular factory, effective strategies for uncertainty management and autonomous systems adaptation as well as the externalization of operational knowledge. This research is part of the Collaborative Research Center (CRC) 1574, which explores these aspects in detail. For an in-depth understanding of specific components, it is referred to other publications by the CRC 1574.

Humans Vs. Service Robots as Social Actors in Persuasion Settings

Our study departs from existing research, which primarily focuses on the benefits of automated social presence in customer service. Instead, we investigate the potential downsides of a service robot’s social presence in marketing persuasion, particularly its influence on consumers’ use of persuasion knowledge. Across five experiments, we aim to: (1) identify factors affecting the social presence of a salesclerk as a persuasion agent (type of service provider, persuasive intent, and robot appearance); and (2) reveal that anthropomorphized service robots in persuasion contexts lead to a decrease in perceived salesclerk sincerity (i.e., social presence effect), ultimately impacting consumer attitudes and behaviors. Additionally, we explore the mechanism behind the social presence effect by examining the moderating role of consumers’ dispositional persuasion knowledge. By applying the persuasion knowledge model to robot-mediated service encounters, this research offers valuable insights into the potential drawbacks of using anthropomorphic robots for marketing persuasion in service relationships.

Improving prediction performance of general protein language model by domain-adaptive pretraining on DNA-binding protein

DNA-protein interactions exert the fundamental structure of many pivotal biological processes, such as DNA replication, transcription, and gene regulation. However, accurate and efficient computational methods for identifying these interactions are still lacking. In this study, we propose a method ESM-DBP through refining the DNA-binding protein sequence repertory and domain-adaptive pretraining based the general protein language model. Our method considers the lacking exploration of general language model for DNA-binding protein domain-specific knowledge, so we screen out 170,264 DNA-binding protein sequences to construct the domain-adaptive language model. Experimental results on four downstream tasks show that ESM-DBP provides a better feature representation of DNA-binding protein compared to the original language model, resulting in improved prediction performance and outperforming the state-of-the-art methods. Moreover, ESM-DBP can still perform well even for those sequences with only a few homologous sequences. ChIP-seq on two predicted cases further support the validity of the proposed method.

Effect of disclosing AI-generated content on prosocial advertising evaluation

With advancements in generative artificial intelligence (AI) technology, there is a growing concern about its ethical implications, transparency, and consumer reactions to AI-generated content. Building on the persuasion knowledge model and algorithm aversion literature, this study explores the effects of AI disclosure in prosocial advertising on consumer attitudes and donation intentions. The findings of Study 1 indicate that the initial disclosure of AI-generated content leads to unfavourable attitudes towards ads, with perceived ad credibility serving as a mediating factor. In Study 2, participants who perceived AI as more human-like rather than machine-like tended to experience a diminished negative impact of AI disclosure. Study 3 also highlights the crucial role of perceived ad credibility in influencing donation intentions following the disclosure of AI-generated content. The theoretical and practical implications of our findings for social marketers and nonprofit organizations are discussed further.

Hierarchical clustering algorithm to generate ontology network

Ontologies have played a crucial role in addressing challenges related to knowledge modeling, artificial intelligence, clustering, and classification. Within the realm of web services, ontologies primarily serve to describe web service interfaces, facilitating the automation of functional processes, particularly to guide the registers in discovering services. Enriching this model with a shared ontology (knowledge base) can improve the efficiency of web services technology by guiding web services processes. This shared ontology plays the role of intermediary between queries and registries, and can be shared between different registries. However, manually constructing ontologies is a labor-intensive and expensive endeavor, necessitating domain expertise. Automatic ontology construction approaches present compelling alternatives, offering advantages in terms of both time and financial costs. In this article, we advocate for the utilization of the Hierarchical Clustering algorithm to generate an ontological network (Service Ontology (SO)) from a semantic web services corpus. Experimental validations are presented to demonstrate the effectiveness of the proposed methodology.

ChatMol: interactive molecular discovery with natural language.

Natural language is poised to become a key medium for human-machine interactions in the era of large language models. In the field of biochemistry, tasks such as property prediction and molecule mining are critically important yet technically challenging. Bridging molecular expressions in natural language and chemical language can significantly enhance the interpretability and ease of these tasks. Moreover, it can integrate chemical knowledge from various sources, leading to a deeper understanding of molecules. Recognizing these advantages, we introduce the concept of conversational molecular design, a novel task that utilizes natural language to describe and edit target molecules. To better accomplish this task, we develop ChatMol, a knowledgeable and versatile generative pretrained model. This model is enhanced by incorporating experimental property information, molecular spatial knowledge, and the associations between natural and chemical languages. Several typical solutions including large language models (e.g. ChatGPT) are evaluated, proving the challenge of conversational molecular design and the effectiveness of our knowledge enhancement approach. Case observations and analysis offer insights and directions for further exploration of natural-language interaction in molecular discovery. Codes and data are provided in https://github.com/Ellenzzn/ChatMol/tree/main.

A Cognitive Model of Language Teacher Fossilization

Expertise in foreign language teaching has been widely discussed in the literature as inherently related to accumulated classroom experience and, hence, extensive practical knowledge. Even though current conceptualizations are increasingly focusing on teachers’ ability to reflect on their performance and continuously develop a repertoire of didactic resources regardless of their career stage, effective human information processing in the profession remains under-researched. This paper sees the concept of expertise from the perspective of cognitive processes, and specifically operations involved in problem-solving. The author presents a theoretical model of teacher fossilization, discussing attention, knowledge, reasoning, and judgment as potential inhibitors of intellectual vitality and growth in individuals. When neglected, these factors might lead experienced teachers to make just as incompetent decisions as novice educators. The model highlights the external and internal forces that cause practitioners to minimize, rather than maximize, their learning opportunities in a workplace. The new concept has implications for teachers’ day-to-day practice as well as for in-service professional development initiatives. It serves as a guiding principle for the successful utilization of one’s cognitive capital and outlines both the purpose and direction of one’s mental effort.

99 An Equity-Oriented Model of Team-Based Care: A Co-Designed Knowledge Mobilization and Model of Care Pilot Intervention in British Columbia, Canada

99 An Equity-Oriented Model of Team-Based Care: A Co-Designed Knowledge Mobilization and Model of Care Pilot Intervention in British Columbia, Canada

Structural Modeling of Food Safety Knowledge, Attitude, and Practices of Food Handlers in Highway Tourist Stop-Over Food Service Establishments in Uganda

The study examines the socio-demographic characteristics, food safety knowledge, attitudes, and practices of food handlers in highway tourist stop-over food service establishments in Uganda. The study further explores the relationship between the above factors using the Structural Equation Model (SEM). Understanding food safety knowledge, attitudes, and practices of food handlers can inform future interventions toward food safety assurance in food service establishments. The study used a cross-sectional design and data was collected using a self-administered structured questionnaire. Data was analyzed using Statistical Package for Social Sciences (SPSS) Version 22 and Analysis of Moment Structures (AMOS). Approximately half of the respondents had adequate knowledge of food safety and good practices. The results from SEM confirmed that food safety knowledge significantly affected practices (β= 0.37 p<0.05), attitude insignificantly, and negatively affected practices (β= -0.04 p>0.05). Only 13.3% of the changes in practices were explained by the knowledge and attitudes of the respondents (R-squared 0.133). The majority of the food handlers were males, had experience of more than 2 years, attended food safety training, and had limited knowledge about Hazard Analysis Critical Control Points (HACCP). The study is the first of its kind in Uganda. The findings suggest that future interventions should look beyond food safety knowledge and attitudes of the food handlers in food safety assurance efforts.

A fine-grained RDF graph model for fuzzy spatiotemporal data

The uncertainty and spatiotemporal dynamics of information necessitate the urgent modeling of fuzzy spatiotemporal knowledge across various applications, with the Resource Description Framework (RDF) serving as a widely recognized data representation model. Existing models suffer from incomplete semantic representation and poor robustness in modeling fuzzy spatiotemporal data, e.g., lack of fuzziness in spatiotemporal semantics; lack of altitude in spatial semantics. Meanwhile, the algebraic framework regarding the model has not been investigated. Thus, in this paper, we first propose a new fine-grained fuzzy spatiotemporal RDF model. This model can represent fine-grained uncertain spatiotemporal semantics that may be associated with any element of a spatiotemporal RDF. We further define its graph algebraic operations. Note that we demonstrate the use of the algebraic operations for fuzzy spatiotemporal RDF querying. Finally, we establish a set of transformation rules for SPARQL query syntax to algebraic operations in fuzzy spatiotemporal RDF. In addition, we used experiments to evaluate the validity and rationality of our model.

Semi-supervised lung adenocarcinoma histopathology image classification based on multi-teacher knowledge distillation.

In this study, we propose a semi-supervised learning scheme using a patch-based deep learning framework to tackle the challenge of high-precision classification of seven lung tumor growth patterns, despite having a small amount of labeled data in whole slide images (WSIs). This scheme aims to enhance generalization ability with limited data and reduce dependence on large amounts of labeled data. It effectively addresses the common challenge of high demand for labeled data in medical image analysis.

Approach. To address these challenges, the study employs a semi-supervised learning approach enhanced by a dynamic confidence threshold mechanism. This mechanism adjusts based on the quantity and quality of pseudo labels generated. This dynamic thresholding mechanism helps avoid the imbalance of pseudo-label categories and the low number of pseudo-labels that may result from a higher fixed threshold. Furthermore, the research introduces a multi-teacher knowledge distillation technique. This technique adaptively weights predictions from multiple teacher models to transfer reliable knowledge and safeguard student models from low-quality teacher predictions.

Main results. The framework underwent rigorous training and evaluation using a dataset of 150 WSIs, each representing one of the seven growth patterns. The experimental results demonstrate that the framework is highly accurate in classifying lung tumor growth patterns in histopathology images. Notably, the performance of the framework is comparable to that of fully supervised models and human pathologists. In addition, the framework's evaluation metrics on a publicly available dataset are higher than those of previous studies, indicating good generalizability.

Significance. This research demonstrates that a semi-supervised learning approach can achieve results comparable to fully supervised models and expert pathologists, thus opening new possibilities for efficient and cost-effective medical images analysis. The implementation of dynamic confidence thresholding and multi-teacher knowledge distillation techniques represents a significant advancement in applying deep learning to complex medical image analysis tasks. This advancement could lead to faster and more accurate diagnoses, ultimately improving patient outcomes and fostering the overall progress of healthcare technology.&#xD.