• Journal of Korea Water Resources Association
• /
• v.55 no.spc1
• /
• pp.1235-1249
• /
• 2022

This study applied the web crawling technique for extracting big data news on water quality accidents in the water supply system and presented the algorithm in a procedural way to obtain accurate water quality accident news. In addition, in the case of a large-scale water quality accident, development patterns such as accident recognition, accident spread, accident response, and accident resolution appear according to the occurrence of an accident. That is, the analysis of the development of water quality accidents through key keywords and sentiment analysis for each stage was carried out in detail based on case studies, and the meanings were analyzed and derived. The proposed methodology was applied to the larval accident period of Incheon Metropolitan City in 2020 and analyzed. As a result, in a situation where the disclosure of information that directly affects consumers, such as water quality accidents, is restricted, the tone of news articles and media reports about water quality accidents with long-term damage in the event of an accident and the degree of consumer pride clearly change over time. could check This suggests the need to prepare consumer-centered policies to increase consumer positivity, although rapid restoration of facilities is very important for the development of water quality accidents from the supplier's point of view.

• Journal of the Korea Convergence Society
• /
• v.10 no.12
• /
• pp.271-280
• /
• 2019

Smart factory, a critical part of digital transformation, enables data-driven decision making using monitoring, analysis and prediction. Predictive maintenance is a key element of smart factory and the need is increasing. The purpose of this study is to analyze the degradation characteristics of a galvanizing kettle for the steel plating process and to predict the remaining useful life(RUL) for predictive maintenance. Correlation analysis, multiple regression, principal component regression were used for analyzing factors of the process. To identify the trend of degradation, a proposed rolling window was used. It was observed the degradation trend was dependent on environmental temperature as well as production factors. It is expected that the proposed method in this study will be an example to identify the trend of degradation of the facility and enable more consistent predictive maintenance.

• Journal of Korean Society of Forest Science
• /
• v.109 no.1
• /
• pp.23-30
• /
• 2020

The use of increased computing power, machine learning, and deep learning techniques have dramatically increased in various sectors. In particular, image detection algorithms are broadly used in forestry and remote sensing areas to identify forest types and tree species. However, in South Korea, machine learning has rarely, if ever, been applied in forestry image detection, especially to classify tree species. This study integrates the application of machine learning and forest image detection; specifically, we compared the ability of two machine learning data collection methods, namely image data captured by forest experts (D1) and web-crawling (D2), to automate the classification of five trees species. In addition, two methods of characterization to train/test the system were investigated. The results indicated a significant difference in classification accuracy between D1 and D2: the classification accuracy of D1 was higher than that of D2. In order to increase the classification accuracy of D2, additional data filtering techniques were required to reduce the noise of uncensored image data.

• Journal of Convergence for Information Technology
• /
• v.11 no.9
• /
• pp.214-220
• /
• 2021

With the rapid spread of smartphones, it has become common to listen to music everywhere, just like background music in life, so it is necessary to create a music database that can make recommendations according to individual circumstances and conditions. This paper proposes a music recommendation model through social media. Since emotions, situations, time of day, weather, etc. are included in hashtags, it is possible to build a social media-based database that reflects the opinions of various people with collective intelligence. We use web crawling to collect and categorize different hashtags from posts with music title hashtags to use real listeners' opinions about music in a database. Data from social media is used to create a music database, and music is classified in a different way from collaborative filtering, which is mainly used by existing music platforms.

• Journal of Digital Convergence
• /
• v.17 no.2
• /
• pp.195-200
• /
• 2019

Nowadays as the use of mobile communication devices such as smart phones and tablets and the use of Computer is expanded, data is being collected exponentially on the Internet. In addition, due to the development of SNS, users can freely communicate with each other and share information in various fields, so various opinions are accumulated in the from of big data. Accordingly, big data analysis techniques are being used to find out the difference between the response of the general public and the response of the media. In this paper, we analyzed the public response in SNS about child allowance and childbirth grant and analyzed the response of the media. Therefore we gathered articles and comments of users which were posted on Twitter for a certain period of time and crawling the news articles and applied sentiment analysis. From these data, we compared the opinion of the public posted on SNS with the response of the media expressed in news articles. As a result, we found that there is a different response to some national policy between the public and the media.

• Journal of Digital Convergence
• /
• v.19 no.5
• /
• pp.1-13
• /
• 2021

The purpose of this research is to analyze how the wage level of new job seekers in the platform labor market affects the period on getting the first job. Recently, the platform gets attention as one of alternatives to solve the increase of unemployment rate. It is important to create quality jobs that we build up a trust between employers and employees in the platform. Previous studies showed that feedback from previous employers is important for solving the information asymmetry problem between those people. However, there is no feedback for new job seekers who have not get the first job. Therefore, we focus on the fact that wages are presented by job seekers rather than employers in the platform, and we will figure out that the low wages of new job seekers may affect the shortening of job-hunting period. For this reason, we use 3,704 job seekers of Freelancer.com. Survival analysis shows that low wages for new job seekers have a significant impact on shortening job-hunting period.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.2
• /
• pp.81-89
• /
• 2022

This study aims to analyze the images of Instagram posts and to draw implcations regarding the exhibition of . This study collects and crawl 24,295 images from Instagram posts as a dataset. We use the Google Cloud Vision API for labeling the images and a total of 212,567 clusters of labels are finally classified into 9 categories using Word2Vec. The categories of museum spaces, photo zone, architecture category are dominant along with people category. In conclusion, visitors curate their experiences and memories of physical places and spaces while they are experiencing with the exhibition. This result reproves the results of previous studies which emphasize a sense of social presence and place making. The convergent approach of art management and art technology used in this study help museum professionals have an insight on big data based visitor research on a practical level.