Earticle

연구목적: 본 연구는 플랜트 건설현장에서 유해·위험요인을 체계적으로 식별하고 평가할 수 있는 딥러닝 기 반 작업 위험성평가 모델을 구축하는 데 목적이 있다. 특히, 경험과 지식이 부족한 근로자도 쉽게 활용할 수 있도록 체계적인 시스템을 설계하여, 기존의 경험기반 평가 방식의 문제를 해결하고자 한다. 연구방법: 국내 플랜트 건설사인 A사의 976,140건의 위험성평가 데이터를 기반으로 데이터 정제 및 전처리를 수행하고, LDA 기반 토픽 모델링과 심층 신경망(DNN)을 활용하여 위험성 평가 모델을 구축하였다. 데이터셋 18,000 개를 사용하여 학습 및 테스트를 진행했으며, 최종 검증은 1,026건의 신규 위험성평가 데이터를 활용하였다. 연구결과: 개발된 모델은 작업조건에 따라 유해·위험 요인과 재해 유형 등을 체계적으로 추천하였고, 추천 결 과의 95.6% 이상이 전문가에 의해 적합하다고 평가되었다. 최종검증을 통해 8개 공종의 위험성평가 1,026개 에 대해 추천된 6,380개 전체가 추천 적합도 80 이상으로 모델 설계 기준을 만족하였고, 전 공종에서 일관된 추천 품질과 높은 적합성을 지님을 입증하였다. 결론: 대규모 비정형 안전 데이터를 정형화하여 데이터 기반 의 체계적인 위험성평가 모델을 구축함으로써, 기존의 경험 및 주관적인 평가 방식에 한계를 보완하였다. AI 기술을 활용을 통해 위험 요소 도출의 일관성과 효율성을 제고하고 위험성 평가의 실효성을 높였다.

Purpose: This study aims to develop a deep learning-based task risk assessment model capable of systematically identifying and evaluating hazardous and risk factors at plant construction sites. The model is specifically designed systematically to be easily utilized by workers with limited experience or knowledge, thereby addressing the issues associated with traditional experience-based assessment methods. Method: A total of 976,140 risk assessment records from domestic construction Company A. were cleansed and pre-processed. A risk assessment model was then constructed using LDA-based topic modeling and a deep neural network(DNN). The model was trained and tested using a dataset of 18,000, with final validation conducted on 1,026 new risk assessments. Result: The developed model systematically recommends hazardous and risky factors and types of accidents based on working conditions, with more than 95.6% of the recommendations evaluated as appropriate by experts. In the final validation, the model recommended 6,380 items for 1,026 assessments across eight work types, all achieving suitability score of above 80, thereby meeting the model design criteria and demonstrating consistent recommendation quality and high applicability across all construction domains. Conclusion: By structuring large-scale unstructured safety data into a data-driven risk assessment model, this study overcomes the limitations of traditional subjective evaluation methods. The application of AI enhances consistency and efficiency in hazard identification, improving the effectiveness of risk assessments.

A wastewater treatment plant is an intricate system with a wealth of information, where the aeration system of the active sludge process is designed to provide oxygen to microorganisms. Owing to the time delay in biochemical reactions, adjustments made by operational staff to the airflow often lead to delayed changes in dissolved oxygen concentration, frequently causing overaeration. The paper introduces a machine learning model that utilizes water quality indicators and air blower indicators to predict current airflow. By leveraging the airflow predicted by this model, the dissolved oxygen concentration for the next hour is successfully maintained within the optimal range of 2 mg/L to 4 mg/L. In the case of airflow prediction, the Transformer model proved more effective than the random forest and long short-term memory models, owing to its self-attention model architecture. In conclusion, the study demonstrates the successful applicability of machine learning models to predict airflow on the promise of maintaining dissolved oxygen stability. These findings present a data-driven approach to guarantee the steady operation of wastewater treatment plants.

The integration of integrated sensing and communication (ISAC) with artificial intelligence (AI)-driven techniques has emerged as a transformative research frontier, attracting significant interest from both academia and industry. As sixth-generation (6G) networks advance to support ultra-reliable, low-latency, and high-capacity applications, machine learning (ML) has become a critical enabler for optimizing ISAC functionalities. Recent advancements in deep learning (DL) and deep reinforcement learning (DRL) have demonstrated immense potential in enhancing ISAC-based systems across diverse domains, including intelligent vehicular networks, autonomous mobility, unmanned aerial vehicles based communications, radar sensing, localization, millimeter wave/terahertz communication, and adaptive beamforming. However, despite these advancements, several challenges persist, such as real-time decision-making under resource constraints, robustness in adversarial environments, and scalability for large-scale deployments. This paper provides a comprehensive review of ML-driven ISAC methodologies, analyzing their impact on system design, computational efficiency, and real-world implementations, while also discussing existing challenges and future research directions to explore how AI can further enhance ISAC’s adaptability, resilience, and performance in next-generation wireless networks. By bridging theoretical advancements with practical implementations, this paper serves as a foundational reference for researchers, engineers, and industry stakeholders, aiming to leverage AI’s full potential in shaping the future of intelligent ISAC systems within the 6G ecosystem.

The need for condition-based maintenance and sustainability management of military assets has become increasingly important due to reductions in military personnel and budget constraints. In this context, research is actively being conducted using sensors that collect environmental data to determine optimal inspection intervals and identify vulnerabilities early. This study proposes an algorithm to dynamically adjust the sampling interval based on the rate of change in measured physical quantities to reduce sensor battery consumption. Various environmental sensing data, such as temperature, humidity, acceleration, and pressure, were collected from a simulation device that mimicked the storage environment of actual guided missiles. The analysis of the collected data revealed that setting the RSME threshold within 5 improved battery lifespans by 11% compared to the original data, while demonstrating that the temperature RMSE was estimated to improve by about 4.76 in adaptive sampling and 5.44 in uniform sampling. This demonstrates that our approach may improve maintenance efficiency by monitoring military assets more effectively and for extended periods.

Advancements in deep learning have enhanced vision-based aggregate analysis. However, further development and studies have encountered challenges, particularly in acquiring large-scale datasets. Data collection is costly and time-consuming, posing a significant challenge in acquiring large datasets required for training neural networks. To address this issue, this study introduces a simulation that efficiently generates the necessary data and labels for training neural networks. We utilized a genetic algorithm (GA) to create optimized lists of aggregates based on the specified values of weight and particle size distribution for the aggregate sample. This enabled sample data collection without conducting sieving tests. Our evaluation of the proposed simulation and GA methodology revealed errors of 1.3% and 2.7 g for aggregate size distribution and weight, respectively. Furthermore, we assessed a segmentation model trained with data from the simulation, achieving a promising preliminary F1 score of 78.18 on the actual aggregate image.

Most of the data warehouse (DW) requirements engineering approaches have not distinguished the early requirements engineering phase from the late requirements engineering phase. There are very few approaches seen in the literature that explicitly model the early & late requirements for a DW. In this paper, we propose an AGDI (Agent-Goal-Decision-Information) model to support the early and late requirements for the development of DWs. Here, the notion of agent refers to the stakeholders of the organization and the dependency among agents refers to the dependencies among stakeholders for fulfilling their organizational goals. The proposed AGDI model also supports three interrelated modeling activities namely, organization modeling, decision modeling and information modeling. Here, early requirements are modeled by performing organization modeling and decision modeling activities, whereas late requirements are modeled by performing information modeling activities. The proposed approach has been illustrated to capture the early and late requirements for the development of a university data warehouse exemplifying our model's ability of supporting its decisional goals by providing decisional information.

This paper proposes how to design and implement data-driven strategies by investigating how a firm can increase its value using data science. Drawing on prior studies on architectural innovation, a behavioral theory of the firm, and the knowledge-based view of the firm as well as the analysis of field observations, the paper shows how data science is abused in dealing with meso-level data while it is underused in using macro-level and alternative data to accomplish machine-human teaming and risk management. The implications help us understand why some firms are better at drawing value from intangibles such as data, data-science capabilities, and routines and how to evaluate such capabilities.

It is crucial to develop effective and efficient big data analytics methods for problem solving in the field of business in order to improve the performance of data analytics and reduce costs and risks in the analysis of customer data. In this study, a big data-driven data analysis system using artificial intelligence techniques is designed to increase the accuracy of big data analytics along with the rapid growth of the field of data science. We present a key direction for big data analysis systems through missing value imputation, outlier detection, feature extraction, utilization of explainable artificial intelligence techniques, and exploratory data analysis. Our objective is not only to develop big data analysis techniques with complex structures of business data but also to bridge the gap between the theoretical ideas in artificial intelligence methods and the analysis of real-world data in the field of business.

The rapid expansion of online apparel retail has increased the demand for accurate size recommendations that minimize returns and enhance customer satisfaction. This study presents a data-driven analysis of a bioelectrical impedance analysis (BIA)–based size recommendation system implemented on a live e-commerce platform. Using anonymized transaction and feedback data from Boastfit.com, the research compares behavioral and perceptual outcomes between BIA-based recommendations and conventional size guides. The BIA group recorded a return rate of 9.6 percent compared with 17.2 percent in the control group, an average satisfaction score above 8 on a ten-point scale, and a repurchase ratio of 79 percent. These results confirm that physiological data–driven personalization improves predictive accuracy, post-purchase satisfaction, and repurchase intention. The findings contribute to next-generation computing and fashion retail analytics by demonstrating how body-composition data can be integrated into intelligent recommendation systems to enhance user trust and sustainable engagement.

The integration of efficient information-based systems characterizes one of the vital priorities of an Italian regional healthcare environment to meet its clinical, organizational, and managerial needs, especially after the COVID-19 pandemic. The current study proposes that the most likely approach to achieve an Italian regional healthcare system is to use a health level-7-based interoperable communication system implemented by an asynchronous communication infrastructure between healthcare sites. The proposed framework is a comprehensive and integrated management information systems (MIS) centered coordination at an Italian regional level that includes all categories of healthcare levels, including the interoperability concerns; that covers most of the needed elements and can work efficiently and effectively in a vast secure area network to safeguard data privacy and confidentiality of a local the patient. Another essential feature of the proposed framework system solution is that it creates interoperability and compatibility that can be simulated from one healthcare institution to another inside the Italian healthcare environment. In that case, joint interoperability and system-compatible messages can interrelate heterogeneous management-based information systems. In response to the COVID-19 situation in Italy, more than 50 different consortiums have submitted proposals directly to the Italian Government. It is believed that the proposed interoperability framework seems to be widely considered an acceptable solution to enhance information and communication technologies (ICT) developments in the healthcare sector in Italy. Furthermore, this study identifies some flaws inside the Italian healthcare sector with possible elucidations for further research.

본 연구는 객관적인 데이터 기반 방법을 통해 인간 중심 디자인 과정을 효과적으로 보완하는 디자인 프로세스를 제시한다. 즉, 주관적 방법에 의한 인간 중심 디자인 프로세스에서 결여되는 객관성이 데이터 기반 접근에 의해 보완되 어 숨겨진 사용자의 니즈를 효과적으로 발견하는 프로세스로 발전될 수 있다. 이에 본 연구에서는 설문조사 데이터 마이닝 분석 과정과 공동 디자인 프로세스가 접목된 인간 중심 디자인 프로세스를 제시하며, 스마트 난방복 사례연구를 통해 이를 검증한다. 설문조사 데이터 마이닝 분석 과정에서는 클러스터링과 의사결정 나무의 두 가지 분석 방법이 사용된다. 클러스터링은 타겟 그룹을 선정하는 기준이 되는 페르소나의 초안을 제시하며, 의사결정 나무는 제품 구매에 중요한 사용자 인식 속성 파악과 사용자 가치 체계를 일차적으로 제안한다. 이후 데이터 분석을 통해 얻어진 광범위한 관점에 대하여 타겟 그룹을 대표하는 사용자가 직접 참여하는 공동 디자인 프로세스가 수행되며 맞춤형 워크북을 이용 하여 신제품에 대한 사용자의 여정맵, 니즈, 아이디어, 가치 체계 등을 체계적으로 도출한다. 본 논문에서 수행한 스마 트 난방복 사례 연구는 제안된 방법론의 적용성을 보여주고 있다.

This research suggests a design process that effectively complements the human-centered design through an objective data-driven approach. The subjective human-centered design process can often lack objectivity and can be supplemented by the data-driven approaches to effectively discover hidden user needs. This research combines the data mining analysis with co-design process and verifies its applicability through the case study on the smart heating jacket. In the data mining process, the clustering can group the users which is the basis for selecting the target groups and the decision tree analysis primarily identifies the important user perception attributes and values. The broad point of view based on the data analysis is modified through the co-design process which is the deeper human-centered design process by using the developed workbook. In the co-design process, the journey maps, needs and pain points, ideas, values for the target user groups are identified and finalized. They can become the basis for starting new product development.

Understanding features of self-repairs is critical to the exploration of interpreters’ underlying self-monitoring mechanism. As most previous taxonomies just made slight amendments on Levelt’s tripartite taxonomy model — which is designed for monolingual speech production, they failed to match features of interpreting output. In order to resolve the defects of previous classifications, the present paper established an interpreting-tailored taxonomy of self-repairs, which categorized repairs into five major types with altogether nine sub-types. Rather than directly referring to Levelt’s model, each category in the new taxonomy is derived from the descriptive study of data collected from ten interpreting trainees, including their interpreting recordings, retrospection and notes. The qualitative analysis on distribution of each repair category also shows that trainees’ adoption of repairs is mainly triggered by competence deficiency rather than the attempt to facilitate communication.

무선센서 네트워크는 센서 데이터베이스 관리 시스템을 통해 보다 효율적으로 개발 및 운용될 수 있다. 센서 데이터베이스 관리 시스템은 무선센서 입력에 대해 선언된 사용자 정의 질의를 처리하기 위해 사용자들에게 익숙한 SQL 유형의 사용자 접속을 지원한다. 무선센서 네트워크상의 전형적 질의 유형은 임의의 스냅 샷 값 검색 이나 오래도록 지속되는 연속 질의 형태를 갖는다. 무선센서 네트워크상에서 질의 처리는 베이스스테이션으로부터 여러 노드들로 질의를 보내는 과정과 여러 노드에서 얻어지는 질의 결과를 베이스스테이션으로 회수하는 과정이 있는데 이러한 질의의 파급이나 베이스스테이션으로의 결과 전송은 많은 에너지 소모를 요구한다. 이 논문은 무선 센서 네트워크상에서 영역 질의를 처리함에 있어 질의 및 결과를 파급시키는데 소모되는 에너지를 절약시켜 주기 위한 분산정보수집(DIG: Distributed Information Gathering)이라고 이름붙인 효율적 색인 방법을 제안한다.

Wireless sensor networks (WSN) can be more reliable and easier to program and use with the help of sensor database management systems (SDMS). SDMS establish a user-friendly SQL-based interface to process declarative user-defined queries over sensor readings from WSN. Typical queries in SDMS are ad-hoc snapshot queries and long-running, continuous queries. In SDMSs queries are flooded to all nodes in the sensor net, and query results are sent back from nodes that have qualified results to a base station. For query flooding to all nodes, and result flooding to the base station, a lot of communication energy consuming is required. This paper suggests an efficient in-network index solution, named Distributed Information Gathering (DIG) to process range queries in a sensor net environment that can save energy by reducing query and result flooding.

With the advent of the Newspace era and the importance of systematic safety management for the unique hazards of the space environment has been highlighted. Conventional safety management approaches have limitations in effectively responding to complex and real-time evolving space risks. This study proposes the necessity of establishing a data-driven and systematic space risk safety management system to protect human lives and assets and ensure the sustainability of space activities. By analyzing the definition and current status of space risks and examining the limitations of existing safety management methods, this research emphasizes the significance of a data-driven risk management system specifically tailored for the space sector. Such a system plays a crucial role in various aspects, including enhancing mission success rates, protecting lives and assets, preventing accidents and providing early warnings, enabling integrated risk analysis across systems, and establishing a robust safety culture. Ultimately, this study aims to contribute to the formulation and implementation of comprehensive space safety management policies for safe and sustainable space development and activities in the Newspace era.

Artificial intelligence algorithms and quantum computing have become core computational methodologies for the applications of artificial intelligence techniques in smart and sustainable cities. In development and sustainability. Artificial intelligence technologies have faced numerous challenges in natural, social, and cultural aspects. The scientific dialogue on artificial intelligence has been significantly extended to understand and cover sustainability as the practice of living enclosed by resources into the distant future. Quantum computing can improve environmental management, applications of artificial intelligence technologies, big data analytics, and utilization of blockchain in smart and sustainable cities. Policy decision-makers may need to consider the introduction of quantum computing technologies for applications of artificial intelligence in their smart cities for urban growth in order to lower barriers to utilizing quantum artificial intelligence.

With the popularization of social media, companies are using it as a new advertising channel to replace traditional advertising channels. As a result, influencer marketing, which utilizes influencers who are highly influential on social media for advertising, is gaining attention. Views are a representative marketing metric, and in this study, we propose a model to predict the number of views of influencer marketing videos using YouTube. To this end, we collected 16,348 influencer marketing videos in the food industry and built a deep learning-based view prediction model using multimodal data such as thumbnails and title text. We also verified the effectiveness of the proposed methodology by performing the same modeling on influencer marketing videos in the tech sector. The results of this study can provide implications for companies and video producers who utilize influencer marketing.

본 연구는 주인-대리인 모델을 적용하여 데이터 산업의 이해관계자인 정부, 개인, 기업의 규제 이슈를 분석해내 는 데 목적이 있다. 데이터 산업은 거대한 딜레마적 상황에 직면해 있다. 데이터 경제의 중요성이 빠르게 부상하고 있으나, 데이터 사용에 대한 국가의 규제로 인해 산업 발전이 저해되는 한편, 데이터의 무분별한 활용으로 인한 개인의 프라이버시 역시 침해받고 있다. 본 연구에서는 기술적 사례연구의 방식을 이용하여 딜레마적 상황에서 각각의 행위자 들의 이해관계에 기반한 규제 이슈를 분석하고, 그에 대응할 수 있는 전략을 제시하였다. 사례분석 결과 첫째, 국내 데이터 산업의 주요 정책행위자는 데이터 회사와 정부이다. 둘째, 데이터 기반 사회에서 가장 우려스러운 두 가지 문제 점은 기업이 빈번하게 개인정보를 침해한다는 것과 국제적 기업의 데이터 독과점 현상이 나타난다는 점이다. 이러한 규제 이슈를 해결하기 위해 본 논문에서는 이에 대한 전략을 다음과 같이 제시하고 있다. 정부는 글로벌 기업의 감독을 위한 국내 대리인제도를 활성화하고 데이터 보호를 증대해야 한다. 기업은 차별적인 규제환경을 해결하고 합법적인 데이터 활용기준을 확장해야 한다. 마지막으로 개인은 능동적인 동의 행태를 구현해야 한다.

This study analyzes the regulatory issues of stakeholders, the firm, the government, and the individual, in the data industry using the principal-agent theory. While the importance of data driven economy is increasing rapidly, policy regulations and restrictions to use data impede the growth of data industry. We applied descriptive case analysis methodology using principal-agent theory. From our analysis, we found several meaningful results. First, key policy actors in data industry are data firms and the government among stakeholders. Second, two major concerns are that firms frequently invade personal privacy and the global companies obtain monopolistic power in data industry. This paper finally suggests policy and strategy in response to regulatory issues. The government should activate the domestic agent system for the supervision of global companies and increase data protection. Companies need to address discriminatory regulatory environments and expand legal data usage standards. Finally, individuals must embody an active behavior of consent.