• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.847-854
• /
• 2022

The aim of this study is to develop a Named Entity Recognition (NER) model to automatically identify construction-related organizations from news articles. This study collected news articles using web crawling technique and construction-related organizations were labeled within a total of 1,000 news articles. The Bidirectional Encoder Representations from Transformers (BERT) model was used to recognize clients, constructors, consultants, engineers, and others. As a pilot experiment of this study, the best average F1 score of NER was 0.692. The result of this study is expected to contribute to the establishment of international business strategies by collecting timely information and analyzing it automatically.

• International conference on construction engineering and project management
• /
• 2011.02a
• /
• pp.516-520
• /
• 2011

Corporate social responsibility (CSR) is of increasing interest to both industry practitioners and academics because of increasing public awareness of environmental, economic and social sustainability. The last decade witnessed a large number of studies on the development of CSR principles and the implementation of CSR in various industries such as the mining industry, oil industry and food industry. However, there is limited number of CSR studies focusing on the construction industry. Considering the critical role the construction industry plays in achieving economic, social and environmental sustainability, it is imperative to extend the current CSR research to the construction sector. This paper provides a critical review of literature related to corporate social responsibility in construction context. A special focus is placed on the current practice adopted by the industry to be socially responsible. A research agenda is set up to tackle this critical issue.

• Asia pacific journal of information systems
• /
• v.32 no.3
• /
• pp.496-513
• /
• 2022

Artificial intelligence or big data technologies can benefit finance companies such as those in the insurance sector. With artificial intelligence, companies can develop better customer segmentation methods and eventually improve the quality of customer relationship management. However, the application of AI-based customer segmentation in the insurance industry seems to have been unsuccessful. Findings from our interviews with sales agents and customer service managers indicate that current customer segmentation in the Korean insurance company relies upon individual agents' heuristic decisions rather than a generalizable data-based method. We propose guidelines for AI-based customer segmentation for the insurance industry, based on the CRISP-DM standard data mining project framework. Our proposed guideline provides new insights for studies on AI-based technology implementation and has practical implications for companies that deploy algorithm-based customer relationship management systems.

• Journal of Korean Society of Forest Science
• /
• v.81 no.2
• /
• pp.124-138
• /
• 1992

This study was carried out to estimate long-term demand functions, and to project consumption of roundwood to the year 2030, using time series data for the period 1970-1990. Especially, the unique features of this study are in the estimation of demand functions for roundwood by species group and by end-use with help of dummy variables. It also, attempts to show how dummy variables can be utilized for improving the estimation result. The result of this study reveals that hardwood roundwood consumption is being substituted by softwood roundwood due to the rapid increase in the relative price of softwood, and this trend is expected to continue in the near future. The consumption of roundwood by mining industry is projected to fall as the coal :mining is expected to decline. The parametric estimates of timber demand function by species group and by end-use indicate that the demand for timber in Korea is more responsive to the performance of domestic economy as a whole, represented by GDP in this study, than to other variables such as own and substitute prices. The effects of population growth and substitute prices could not be determined.

• Journal of the Korean Wood Science and Technology
• /
• v.10 no.3
• /
• pp.194-195
• /
• 1982

Recommendaton: These are the highlights of the findings of the Timber Consumption Survey carried out by the Project in 1966, and covering consumption for the period from 1961 to 1965. The survey was oriented towards consumption for structural, commercial and industrial purposes and existing estimates for local (village-level) consumption as fuel and the like were adopted. A full report on the survey was submitted to the Bureau of Forestry in 1966. Long-term Trends: After allowance for anticipated population increase, this ten year's increase in industrial wood consumption represents a gain of about 30% in per capita consumption (from 0.0913 cu.m. per capita to 0.118 cu.m. per capita). This is only about half the expected general economic growth of about 75% (7% per annum). It is therefore likely (a) that the 1975 estimate is conservative, (b) that the consumption demand beyond 1975 may be expected to build up at a greatly increased rate. Estimated income elasticity coefficients are high, and with expected ir,creases in prosperity and population, the consumption is expected to rise to 10 million cu. meters by the year 2,000. Consumption Pattern: The breakdown of industrial consumption (1965) is given in Table 4-2, showing sawnwood consumption as the most important in 1965. The upward trend in all sectors over the 1961-65 period is expected to continue. The general consumption pattern is expected to change through 1975 with a sharp increase in the relative importance of pulp products (to 30% of total consumption) offset by declining relative importance of sawlogs. The following recommendations follow from the study: (i) Industrial forests. - A programme of establishment of consolidated industrial forests should be initiated as a matter of urgency. (ii) Fuelwood forests - Properly sited, protected and managed fuelwood forest, worked on a 20-year rotation, should be established as a nation wide basis. (iii) Hardwood utilization - Detailed investigations are required into the use of indigenous hardwoods for the pulp, particle board and hardboard industries. (iv) Mining timber - Preservation treatment of all mining timber should be enforced by law. (v) Sawmills - Licencing restrictions should be enforced to reduce the number of small, inefficient sawmills. b. Extension work should be undertaken bv government to improve sawmilling practices.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.11
• /
• pp.453-462
• /
• 2016

Since the Internet provides a way of expressing and sharing Internet users' mindsets, corporate marketers want to acquire measurable and actionable insights from web data. In the past, companies used to analyze the attitude, satisfaction, and loyalty of consumers toward their brands using survey data, whereas nowadays this is done using the big data extracted from Social Network Services. In this study, we propose a framework for clustering brand names using the social metrics gathered on social media. We also conduct a case study of the automobile industry to verify the feasibility of the proposed framework. We calculate the brand name distance for each pair of brand names based on the total number of times that they are mentioned together. These distances are used to project the brand name onto a 3-dimensional space using multidimensional scaling. After the projection, we found the clusters of brand names and identified the characteristics of each cluster. Furthermore, we concluded this paper with a discussion of the limitations and future directions of this research.

• Safety and Health at Work
• /
• v.4 no.4
• /
• pp.201-209
• /
• 2013

Background: The study was conducted to investigate the current status of the occupational health and safety management system (OHSMS) in the construction industry and the effect of OHSMS on accident rates. Differences of awareness levels on safety issues among site general managers and occupational health and safety (OHS) managers are identified through surveys. Methods: The accident rates for the OHSMS-certified construction companies from 2006 to 2011, when the construction OHSMS became widely available, were analyzed to understand the effect of OHSMS on the work-related injury rates in the construction industry. The Korea Occupational Safety and Health Agency 18001 is the certification to these companies performing OHSMS in South Korea. The questionnaire was created to analyze the differences of OHSMS awareness between site general managers and OHS managers of construction companies. Results: The implementation of OHSMS among the top 100 construction companies in South Korea shows that the accident rate decreased by 67% and the fatal accident rate decreased by 10.3% during the period from 2006 to 2011. The survey in this study shows different OHSMS awareness levels between site general managers and OHS managers. The differences were motivation for developing OHSMS, external support needed for implementing OHSMS, problems and effectiveness of implementing OHSMS. Conclusion: Both work-related accident and fatal accident rates were found to be significantly reduced by implementing OHSMS in this study. The differences of OHSMS awareness between site general managers and OHS managers were identified through a survey. The effect of these differences on safety and other benefits warrants further research with proper data collection.

• The Journal of Society for e-Business Studies
• /
• v.25 no.1
• /
• pp.203-214
• /
• 2020

It is necessary to classify technical documents such as patents, R&D project reports in order to understand the trends of technology convergence and interdisciplinary joint research, technology development and so on. Text mining techniques have been mainly used to classify these technical documents. However, in the case of classifying technical documents by text mining algorithms, there is a disadvantage that the features representing technical documents must be directly extracted. In this study, we propose a BERT-based document classification model to automatically extract document features from text information of national R&D projects and to classify them. Then, we verify the applicability and performance of the proposed model for classifying documents.

• Science of Emotion and Sensibility
• /
• v.12 no.2
• /
• pp.233-242
• /
• 2009

Recently the studies on user-centered design and form-development have become issues of general interest as the key methods for successful design. For form analysis on user it is important needs that an integrated approach of existing methods and development of expert tool for designer. Moreover analysis methods and tools have to meet with the designers needs of visual result, clear direction, concrete formative factor, user's emotional response and designer-friendly interface. This study proposed the main concepts of form analysis tool based on the user's emotional response ; integrated management, variables set-up, visual result of analysis, in-depth analysis with data mining and correlation, and reinforcement of user-centered analysis. Specific analysis tool consists of 5 functions: Project Management, Analysis Frame Set-up, Data Input-output, Basic Analysis, and In-depth Analysis. The feasibility of proposed tool was verified by a case study of mobile phone design in under-graduate class.

• Journal of Korean Society on Water Environment
• /
• v.35 no.3
• /
• pp.257-270
• /
• 2019

Followingthe Four Rivers Project, cyanobacterial blooms have been frequently observed in the upstream of the installed weirs. The aim of this study was to characterize the major environmental factors that are associated with the cyanobacteria dominance in Baekje Weir (Geum River) and Juksan Weir (Youngsan River) based on intensive experiments and systematic data mining methods. The factors related to the cyanobacteria dominance include7-days cumulative rainfall (APRCP7), 7-days averaged flow (Q7day), water temperature (Temp), stratification strength (${\Delta}T$), electronic conductivity (EC), DO, pH, $NO_3-N$, $NH_3-N$, TN, TP, $PO_4-P$, Chl-a, Fe, BOD, COD, TOC, and $SiO_2$. The most highly correlatedfactors to the dominant cyanobacteria were found to be EC, Temp, Q7day, $PO_4-P$ in theBaekje Weir. On the other hand, those dominant in the Juksan Weir were ${\Delta}T$, TOC, Temp, EC and TN. The EC showed a strong correlation with cyanobacteria dominance in both weirs because a high EC represents a persisted low flow condition. The cyanobacteria dominance was as high as 56 % when the EC was equal or greater than $418{\mu}S/cm$ in Baekje Weir. It was as high as 63% when the ${\Delta}T{\geq}2.1^{\circ}C$ in the Juksan Weir. However, nutrients showed a minor correlation with cyanobacteria dominance in both weirs. The results suggest that the cyanobacteria dominate in astate where the water flow rate is low, water temperature is high and thermal stratification is strengthened. Therefore, the improvement of flow regimes is the most important to prevent persistent thermal stratification and formation of cyanobacteria bloom in theBaekje and JuksanWeirs.