모빌리티 예측은 단순한 통행 경로 예측을 넘어, 사회 전반의 효율성 및 안전성 향상을 위한 핵심 데이터를 제공한다는 점에서 중 요하다. 기존의 예측 기법은 시공간적 규칙성과 개인 이동 패턴의 통계적 특성 분석에 주로 의존하였으며, 최근 딥러닝 기반의 시공간 모델링을 통해 예측 성능이 향상되었다. 그러나 여전히 개인 통행의 단기·장기적 시공간 의존성 및 복잡한 패턴을 처리하는 데 한계가 존재한다. 이를 극복하기 위해, 본 연구는 대규모 사전 학습된 거대 언어 모델(Large Language Model; LLM)을 도입하여, 개인 속성뿐 만 아니라 실제 통행 데이터를 반영한 객체 단위 통행 생성 프레임워크를 제안한다. LLM 기반(ChatGPT-4o) 객체 단위 통행 생성 프레 임워크는 (1) 개인 모빌리티 패턴 학습, (2) 통행 생성의 두 단계로 이루어진다. 이후 한국교통연구원의 개인통행 실태조사(2021) 데이 터를 이용하여 프레임워크의 통행 생성 성능을 확인하였다. 통행 시작·출발 시간 분포, 출발·도착지 장소 유형, 통행목적, 이용 교통수 단의 정확도를 확인한 결과, 대부분 항목에서 70% 이상의 정확도를 보였다. 하지만 통행목적은 13개의 목적 중 하나를 예측해야 하기 에 정확도가 다른 항목에 비해 약 40%로 낮게 나타났다. 본 연구는 통행 생성 프레임워크를 설계하고, 이에 맞춰 입력 데이터를 가공 및 프롬프트 엔지니어링을 수행함으로써 LLM 기반 통행 생성 기술의 가능성을 확인하였다. 향후 프레임워크의 예측 성능 검증 및 개 선을 위한 추가 연구가 필요하며, 날씨, 대규모 행사 등과 같은 외부 요인들을 고려하면 더욱 정교하고 현실적인 통행일지를 생성할 수 있을 것이다.

최근 인공지능 분야에서 가장 활발히 연구되고 있는 거대 언어 모델은 교육에 대한 응용 가능성을 보 이며, 교육학의 거의 모든 분야에서 그 활용 방안이 연구되고 있다. 이러한 연구는 공학 교육에서도 주목 받고 있다. 그러나 구체적인 활용 분야와 방법에 대해서는 아직 많은 연구가 필요한 상황이다. 특히, 거대 언어 모델을 이용한 교육과정 설계와 개선에 대한 연구는 인공지능 공학과 교육학 두 분야에서 모두 중요한 연구 과제로 부각되고 있다. 이러한 응용 필요성에 대한 예시이자 전략으로써, 본 연구는 OpenAI에서 발표한 최신 거대 언어 모델인 ChatGPT-4o를 이용하여 한국과학기술원(KAIST) 공과대학 학부 전공 과 목과 S전자 DS부문(반도체사업부) 직무 사이의 연관성을 분석하고, 그 결과를 기반으로 대학과 기업체 양측에 반도체 산업 인력 양성과 채용에 대한 실질적인 응용 전략을 제안한다. 이를 위해 KAIST 공과대 학 학부과정에 개설된 모든 전공 과목과 S전자 DS부문(반도체사업부)의 직무기술서를 ChatGPT-4o에 학습시켜 각 과목이 특정 제품군, 직무와 가지는 연관성을 특정 범위와 기준에 의거하여 정량화된 점수로 평가했다. 또한, 각각의 직무, 전공, 과목별로 확보한 데이터를 기초적인 통계 분석을 통해 평가했으며, 구직자와 구인자의 활용 가능성에 초점을 두고 특정 전공의 각 직무별 연관성과 특정 직무의 각 전공별 연관성, 그리고 특정 직무 및 전공의 반도체 제품군별 연관성 등 다양한 조건에서 분석을 진행하였다. 또 한 본 전략에 대한 반도체 산업 실무자 견해를 수집하여 실제 전략으로의 활용 가능성을 검증하였다. 분 석 결과, 간단한 질문과 분석만으로도 전공, 교과목별로 유의미한 직무 연관성의 차이를 확인했다. 이러한 결과를 바탕으로 본 연구는 대학 교육과정의 개선과 기업 채용 및 양성 과정에서의 응용 전략을 제시한 다. 이 연구는 대학과 산업 간의 협력을 통해 인적자원 개발과 채용 효율성 증대에 기여할 것으로 기대한 다. 또한, 후속 연구로 구직자와 구인자, 교수자 등 본 연구의 효과를 확인할 수 있는 집단을 대상으로 한 대규모 설문조사 및 전문가그룹 대상 질적연구 등을 제안하여 실제 활용도와의 비교 분석 연구를 제안 한다. 결론적으로, 본 연구는 거대 언어 모델을 활용하여 필요한 인재를 양성하기 위한 교육 과정 설계의 구체적인 응용 가능성을 제시함으로써, 인공지능을 이용한 교육 분야에 대한 기여 방안을 모색한다.

The Nuclear Export and Import Control System (NEPS) is currently in operation for nuclear export and import control. To ensure consistent and efficient control, various computational systems are either already in place or being developed. With numerous scattered systems, it becomes crucial to integrate the databases from each to maximize their utility. In order to effectively utilize these scattered computer systems, it is necessary to integrate the databases of each system and develop an associated search system that can be used for integrated databases, so we investigated and analyzed the AI language model that can be applied to the associated search system. Language Models (LM) are primarily divided into two categories: understanding and generative. Understanding Language Models aim to precisely comprehend and analyze the provided text’s meaning. They consider the text’s bidirectional context to understand its deeper implications and are used in tasks such as text classification, sentiment analysis, question answering, and named entity recognition. In contrast, Generative Language Models focus on generating new text based on the given context. They produce new textual content continuously and are beneficial for text generation, machine translation, sentence completion, and storytelling. Given that the primary purpose of our associated search system is to comprehend user sentences or queries accurately, understanding language models are deemed more suitable. Among the understanding language models, we examined BERT and its derivatives, RoBERTa and DeBERTa. BERT (Bidirectional Encoder Representations from Transformers) uses a Bidirectional Transformer Encoder to understand the sentence context and engages in pre-training by predicting ‘MASKED’ segments. RoBERTa (A Robustly Optimized BERT Pre-training Approach) enhances BERT by optimizing its training methods and data processing. Although its core architecture is similar to BERT, it incorporates improvements such as eliminating the NSP (Next Sentence Prediction) task, introducing dynamic masking techniques, and refining training data volume, methodologies, and hyperparameters. DeBERTa (Decoding-enhanced BERT with disentangled attention) introduces a disentangled attention mechanism to the BERT architecture, calculating the relative importance score between word pairs to distribute attention more effectively and improve performance. In analyzing the three models, RoBERTa and DeBERTa demonstrated superior performance compared to BERT. However, considering factors like the acquisition and processing of training data, training time, and associated costs, these superior models may require additional efforts and resources. It’s therefore crucial to select a language model by evaluating the economic implications, objectives, training strategies, performance-assessing datasets, and hardware environments. Additionally, it was noted that by fine-tuning with methods from RoBERTa or DeBERTa based on pre-trained BERT models, the training speed could be significantly improved.

This study was conducted to develop a model for predicting the growth of kimchi cabbage using image data and environmental data. Kimchi cabbages of the ‘Cheongmyeong Gaual’ variety were planted three times on July 11th, July 19th, and July 27th at a test field located at Pyeongchang-gun, Gangwon-do (37°37′ N 128°32′ E, 510 elevation), and data on growth, images, and environmental conditions were collected until September 12th. To select key factors for the kimchi cabbage growth prediction model, a correlation analysis was conducted using the collected growth data and meteorological data. The correlation coefficient between fresh weight and growth degree days (GDD) and between fresh weight and integrated solar radiation showed a high correlation coefficient of 0.88. Additionally, fresh weight had significant correlations with height and leaf area of kimchi cabbages, with correlation coefficients of 0.78 and 0.79, respectively. Canopy coverage was selected from the image data and GDD was selected from the environmental data based on references from previous researches. A prediction model for kimchi cabbage of biomass, leaf count, and leaf area was developed by combining GDD, canopy coverage and growth data. Single-factor models, including quadratic, sigmoid, and logistic models, were created and the sigmoid prediction model showed the best explanatory power according to the evaluation results. Developing a multi-factor growth prediction model by combining GDD and canopy coverage resulted in improved determination coefficients of 0.9, 0.95, and 0.89 for biomass, leaf count, and leaf area, respectively, compared to single-factor prediction models. To validate the developed model, validation was conducted and the determination coefficient between measured and predicted fresh weight was 0.91, with an RMSE of 134.2 g, indicating high prediction accuracy. In the past, kimchi cabbage growth prediction was often based on meteorological or image data, which resulted in low predictive accuracy due to the inability to reflect on-site conditions or the heading up of kimchi cabbage. Combining these two prediction methods is expected to enhance the accuracy of crop yield predictions by compensating for the weaknesses of each observation method.

PURPOSES : In this study, a model was developed to estimate the concentrations of particulate matter (PM2.5 and PM10) in expressway tunnel sections. METHODS : A statistical model was constructed by collecting data on particulate matter (PM2.5 and PM10), weather, environment, and traffic volume in the tunnel section. The model was developed after accurately analyzing the factors influencing the PM concentration. RESULTS : A machine learning-based PM concentration estimation model was developed. Three models, namely linear regression, convolutional neural network, and random forest models, were compared, and the random forest model was proposed as the best model. CONCLUSIONS : The evaluation revealed that the random forest model displayed the least error in the concentration estimation model for (PM2.5 and PM10) in all tunnel section cases. In addition, a practical application plan for the model developed in this study is proposed.

A machine learning-based algorithms have used for constructing species distribution models (SDMs), but their performances depend on the selection of backgrounds. This study attempted to develop a noble method for selecting backgrounds in machine-learning SDMs. Two machine-learning based SDMs (MaxEnt, and Random Forest) were employed with an example species (Spodoptera litura), and different background selection methods (random sampling, biased sampling, and ensemble sampling by using CLIMEX) were tested with multiple performance metrics (TSS, Kappa, F1-score). As a result, the model with ensemble sampling predicted the widest occurrence areas with the highest performance, suggesting the potential application of the developed method for enhancing a machine-learning SDM.

Recently, as the possibility of unexpected outbreaks of alien insects has increased due to climate change such as global warming, the importance of early control through rapid and accurate spread of exotic forest pest and change prediction diagnosis is required. This study summarizes and reports the followings: the establishment of monitoring strategy for exotic insects by the investigation of species distribution range through field surveys and others, the development of new diagnostic technique through microstructures and life-cycle, the dispersal of exotic insects, and ecological impact assessment using ecological methods and with the expansion of exotic insects and development of ecosystem impact prediction model.

In the current era of sustainable development, economic, social, and environmental changes are interrelated, and social inclusion and environmental sustainability are our shared goals. In response, social and environmental values have become important considerations for the success of an enterprise, placing an increased emphasis on the interests of all stakeholders. This trend in the governance of enterprise fueled the emergence of a new organizational form: the certified B Corporation (B Corp), a social enterprise certified by B Lab as an enterprise that creates value for non-shareholding stakeholders, including employees, customers, the local community, and the environment. With their positive social and environmental impacts, B Corps have become increasingly recognized as instrumental in the achievement of the UN Sustainable Development Goals, and literature on B Corps has increased. However, empirical research on the role of B Corps is still lacking.

Due to COVID-19, changes in consumption trends are taking place in the distribution sector, such as an increase in non-face-to-face consumption and a rapid growth in the online shopping market. However, it is difficult for small and medium-sized export sellers to obtain forecast information on the export market by country, compared to large distributors who can easily build a global sales network. This study is about the prediction of export amount and export volume by country and item for market information analysis of small and medium export sellers. A prediction model was developed using Lasso, XGBoost, and MLP models based on supervised learning and deep learning, and export trends for clothing, cosmetics, and household electronic devices were predicted for Korea's major export countries, the United States, China, and Vietnam. As a result of the prediction, the performance of MAE and RMSE for the Lasso model was excellent, and based on the development results, a market analysis system for small and medium sellers was developed.

PURPOSES : This study aims to analyze the impact of demand risk on two public-private partnership (PPP) projects, namely BTO and BTO-a. The main aspects covered in this study are: i) identification of key risk issues considering the structure of PPP projects, and ii) game theory-oriented scenario building and simulation of demand risk allocation from participants’ perspectives. METHODS : Using the institutional analysis and development (hereafter IAD) framework, a hypothetical structure is formulated to examine the interactions of demand risk. It develops a series of demand risk allocation models for PPP projects (i.e., BTO and BTO-a). The risk structures from the IAD step are the demand risk allocation issues. Using game theory-oriented simulation, this study evaluates demand risk based on scenario building. RESULTS : First, this study highlights the imbalanced rate problems of returns between the BTO and BTO-a projects proposed by the market. This may lead to improvement measures geared towards problematic methods for determining the rate of return among domestic PPP projects. Second, compared with the BTO type, this study expects that the BTO-a type may exhibit more effectiveness, which can increase the probability of project success in both the public and private sectors. Third, judging from game-theory-oriented approaches, this study confirms the function of the BTO-a as a method to adjust moral hazard in the private sector. CONCLUSIONS : Government management standards for BTO-a projects were derived based on the simulation results. It is necessary to select an appropriate project method based on rationality by balancing the IRR for each project method. Legal regulations should be applied separately to each part of the government guarantee. In addition, this study emphasizes that the introduction of ex-post value-for-money (VFM) analysis is essential for the efficient management of government expenses.

The origin and evolution of Chinese characters (hanzi, 漢字), the reasons for its stability and longevity, and its future are some core issues in the study of Chinese characters. We have proposed three research models to tackle these problems: (1). Taking advantage of the logographic nature of Chinese characters, we have used a mathematical model to show that the Chinese writing should have already existed no later than 2100 BCE. (2). We have adopted the “funnel model” of protein folding in biochemistry to illustrate the landscape at the beginning of Chinese writing and how it evolved into a stable writing system. (3). We have proposed an ecological model for studying the past and future of Chinese characters. Based on these models, together with systematic archaeological study of pottery inscriptions and DNA analysis of human skeletons unearthed from various neolithic cultural sites, this article discuss specific issues related to the genesis, the longevity and the future growth of Chinese characters in the context of ecological model of Chinese characters. Particularly, how Chinese characters can be prepared to respond to future challeneges in a world of globalization and dataism.

PURPOSES : The main purpose of this study is to identify directions for improvement of triangular islands installation warrants through analysis of the characteristics of crashes and severity with and without triangular islands on intersections. METHODS : The data was collected by referring to the literature and analyzed using statistical analysis tools. First, an independence test analyzed whether statistically significant differences existed between crashes depending on the installation of triangular islands. As a result of the analysis, individual prediction models were developed for cases with significant differences. In addition, each crash factor was derived by comparison with each model. RESULTS : Significant differences appeared in the "crash frequency of serious or fatal" and "crash severity" owing to the installation of triangular islands. As a result of comparing crash factors through the individual models, it was derived that the differences were dependent on the installation of the triangular islands. CONCLUSIONS : As a result of reviewing previous studies, it is found that improving the installation warrants of triangular islands is reasonable. Through this study, the need to consider the volume and composition ratio of right-turn vehicles when installing a triangular island was also derived; these results also need to be referred to when improving the triangular island installation warrants.

This study explored the usefulness and implications of the Bayesian hyperparameter optimization in developing species distribution models (SDMs). A variety of machine learning (ML) algorithms, namely, support vector machine (SVM), random forest (RF), boosted regression tree (BRT), XGBoost (XGB), and Multilayer perceptron (MLP) were used for predicting the occurrence of four benthic macroinvertebrate species. The Bayesian optimization method successfully tuned model hyperparameters, with all ML models resulting an area under the curve (AUC) > 0.7. Also, hyperparameter search ranges that generally clustered around the optimal values suggest the efficiency of the Bayesian optimization in finding optimal sets of hyperparameters. Tree based ensemble algorithms (BRT, RF, and XGB) tended to show higher performances than SVM and MLP. Important hyperparameters and optimal values differed by species and ML model, indicating the necessity of hyperparameter tuning for improving individual model performances. The optimization results demonstrate that for all macroinvertebrate species SVM and RF required fewer numbers of trials until obtaining optimal hyperparameter sets, leading to reduced computational cost compared to other ML algorithms. The results of this study suggest that the Bayesian optimization is an efficient method for hyperparameter optimization of machine learning algorithms.

본 연구에서는 농산물에서 오염 가능성이 있는 병원성 식중독 균 L. monocytogenes에 대해 신선편의 샐러드, 파인애플, 냉동망고에서 예측 모델을 개발하고, 본 연구에서 개발된 예측 모델을 다른 제품에서 적용 여부를 검증하였다. 시료에 L. monocytogenes를 접종하여 각각의 저장 온도에 보관 시 샐러드는 13oC, 파인애플은 10oC 이상에서 성장하였으며, 두 식품 중 파인애플에서 L. monocytogenes 가 더 빠르게 성장하는 것으로 확인 되었다. 또한, 냉동망 고에 접종한 L. monocytogenes는 -2, -10, -18oC의 저장온 도에서 온도가 낮아질수록 delta 값이 커지며 생존력이 높아지는 양상을 보였다. 본 실험 검증을 통해 같은 신선편 의 과일, 채소 식품 그룹에 속하더라도 식품 각각의 특성에 따라 L. monocytogenes의 성장 패턴은 일정하지 않으며 각기 다른 행동 패턴을 보이는 것으로 확인하였다. 신 선편의 샐러드 및 절단된 과일류는 냉장유통 되며 추가세 척 없이 소비되는 제품 특성상 공정과정에서 L. monocytogenes에 의한 오염이 일어나지 않도록 위생관리 에 주의하고 유통과정에서 온도 남용이 되지 않도록 유통 온도 관리에도 유의해야할 것으로 사료된다.