• International conference on construction engineering and project management
• /
• 2005.10a
• /
• pp.587-591
• /
• 2005

The case of laboratory-acquired SARS Corona virus infection in Taiwan has revealed a number of weaknesses in management, construction, and oversight of laboratories. Also, with the increased demands for bio-safety laboratory, there is an urgent need to develop a uniform and comprehensive guidance for architects and construction engineers in the preparation of design and construction. This research investigates the key elements for designers, engineers, and potential owners in biosafety laboratory design and construction. It defines key elements and determines major relationships and standards that should be adhered to when developing site layout. In addition to layout planning and design guidance of biosafety laboratory, this research also interviews the perspective of architects and survey the state-of-the-art technology in Taiwan. It represents the portraits by site investigation. The purpose of the research is to provide guideline of design and avoid potential future conflict to ensure the critical continuity of functions.

• International conference on construction engineering and project management
• /
• 2005.10a
• /
• pp.480-485
• /
• 2005

This paper reviews the literature on safety culture focusing particularly on research carried out from 1998 onwards. The term 'safety culture' is clarified as it is typically applied to organizations, to safety and particularly to construction safety. Some clarifications in terms of levels of aggregation, positive safety culture and safety performance are provided by presenting appropriate empirical evidences and their theoretical developments. Safety culture is a subset of organizational culture that is thought to influence employees' attitudes and behavior in relation to an organization's ongoing health and safety performance. Implications for future research in the area are addressed, as safety culture has in recent years become the focus of much attention in all industries, and in the construction industry in particular.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2021.06a
• /
• pp.97-100
• /
• 2021

Light field (LF) superpixel segmentation aims to group the similar pixels not only in the single image but also in the other views to improve the computational efficiency of further applications like object detection and pattern recognition. Among the state-of-the-art methods, there is an approach to segment the LF images while enforcing the view consistency. However, it leaves too much noise and inaccuracy in the shape of superpixels. In this paper, we modify the process of the clustering step. Experimental results demonstrate that our proposed method outperforms the existing method in terms of view-consistency.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.388-391
• /
• 2019

과편향 뉴스 판별(hyperpartisan news detection)은 뉴스 기사가 특정 인물 또는 정당에 편향되었는지 판단하는 task이다. 이를 위해 feature-based ELMo + CNN 모델이 제안되었으나, 이는 문서 임베딩이 아닌 단어 임베딩의 평균을 사용한다는 한계가 존재한다. 따라서 본 논문에서는 feature-based 접근법을 따르며 Sentence-BERT(SentBERT)의 문서 임베딩을 이용한 feature-based SentBERT 기반의 과편향 뉴스 판별 모델을 제안한다. 제안 모델의 효과를 입증하기 위해 ELMO, BERT, SBERT와 CNN, BiLSTM을 적용한 비교 실험을 진행하였고, 기존 state-of-the-art 모델보다 f1-score 기준 1.3%p 높은 성능을 보였다.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.297-299
• /
• 2021

Text classification is one of the popular tasks in Natural Language Processing (NLP) used to classify text or document applications such as sentiment analysis and email filtering. Nowadays, state-of-the-art (SOTA) Machine Learning (ML) and Deep Learning (DL) algorithms are the core engine used to perform these classification tasks with high accuracy, and they show satisfying results. This paper conducts a benchmarking performance's analysis of multiple SOTA algorithms on the first known labeled Korean voice phishing dataset called KorCCVi. Experimental results reveal performed on a test set of 366 samples reveal which algorithm performs the best considering the training time and metrics such as accuracy and F1 score.

• Ocean Systems Engineering
• /
• v.13 no.1
• /
• pp.1-20
• /
• 2023

An iterative focused wave generation method is developed and implemented in a local analytic based Navier-Stokes solver. This wave generation method is designed to reproduce the target focused wave by matching the target amplitude spectrum and phase angle. A 4-waves decomposition scheme is utilized to obtain the linearised component of the output wave. A model test studying the interaction between different focused waves and a fixed cylinder is selected as the target for the wave generation approach. The numerical wave elevations and dynamic pressure on the cylinder are compared with the experimental measurement and other state-of-the-art numerical methods' results. The overall results prove that the iterative adjustment method is able to optimize the focused wave generated by a CFD approach.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.41-43
• /
• 2015

Urban Informatics, the application of data science methodologies to the urban development and planning domain, has been increasingly adopted to improve the management and efficiency of cities. This paper introduces state of the art use cases in major cities including New York, London, Seoul and Amsterdam. It also introduces recent advances in using Big Data by multi-lateral institutions for poverty reduction, and startups utilizing open data initiatives to create new value and insights. Preliminary research performed on using Seoul's open data such as building permit data and health code violations are also introduced to demonstrate opportunities in this relatively new but promising area of research.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.21-22
• /
• 2015

Productivity has been a wide-ranging challenge for the construction industry, both in Australia and globally. Particularly in Western Australia's construction and resources sectors, continuously low productivity will potentially discourage future investments. The emergence of the global marketplace necessitates that the supply chain needs to focus on the concept of the holistic efficiency. The isolated geographical position of Australia only exacerbates this phenomenon. In recent years, Building Information Modelling (BIM) has been suggested as an efficient way to help productivity improve and information management throughout supply chains. This keynote talk will focus on discussing ways of implementing BIM to enhance site productivity focusing on Western Australia's construction projects. It will show new trends of its applications to accomplish an innovative way in construction project management. The talk will also give an insightful summary of integrated methods with state-of-the-art technologies backboned by the BIM cases from construction and oil and gas industry projects.

• Annual Conference of KIPS
• /
• 2023.11a
• /
• pp.628-631
• /
• 2023

We present a new approach to evaluate the generated texts by Large Language Models (LLMs) for meme classification. Analyzing an image with embedded texts, i.e. meme, is challenging, even for existing state-of-the-art computer vision models. By leveraging large image-to-text models, we can extract image descriptions that can be used in other tasks, such as classification. In our methodology, we first generate image captions using BLIP-2 models. Using these captions, we use GPT-4 to evaluate the relationship between the caption and the meme text. The results show that OPT_6.7B provides a better rating than other LLMs, suggesting that the proposed method has a potential for meme classification.

• International journal of advanced smart convergence
• /
• v.13 no.1
• /
• pp.92-98
• /
• 2024

The current software industry has an emerging issue with natural language-based requirement specifications. However, the accuracy of such requirement analysis remains a concern. It is noted that most errors still occur at the requirement specification stage. Defining and analyzing requirements based on natural language has become necessary. To address this issue, the linguistic theories of Chomsky and Fillmore are applied to the analysis of natural language-based requirements. This involves identifying the semantics of morphemes and nouns. Consequently, a mechanism was proposed for extracting object state designs and automatically generating code templates. Building on this mechanism, I suggest generating natural language-based comic images. Utilizing state diagrams, I apply changes to the states of comic characters (protagonists) and extract variations in their expressions. This introduces a novel approach to comic image generation. I anticipate highly productive comic creation by applying software processes to Cartoon ART.