• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.3
• /
• pp.879-886
• /
• 2018

The basic search engine on Twitter shows not only tweets that contain search keywords, but also all articles written by users with nicknames containing search keywords. Since the tweets unrelated to the search keyword are exposed as search results, it is inconvenient to many users who want to search only tweets that include the keyword. To solve this inconvenience, this study improved the search function of Twitter by developing an algorithm that searches only tweets that contain search keywords. The improved functionality is implemented as a Web service using ASP.NET MVC5 and is available to many users. We used a powerful collection method in C# to retrieve the results of an object, and it was also possible to output them according to the number of 'retweets' or 'favorites'. If the number of retrieved numbers is less than a given number, we also added an exclusion filter function. Thus, sorting search results by the number of retweets or favorites, user can quickly search for opinions that are of interest to many users. It is expected that many users and data analysts will find the developed function convenient to search on Twitter.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.3
• /
• pp.57-67
• /
• 2018

The ensemble is a unified approach used for getting better performance by using multiple algorithms in machine learning. In this paper, we introduce boosting and bagging, which have been widely used in ensemble techniques, and design a method using support vector regression, radial basis function network, Gaussian process, and multilayer perceptron. In addition, our experiment was performed by adding a recurrent neural network and MOHID numerical model. The drifter data used for our experimental verification consist of 683 observations in seven regions. The performance of our ensemble technique is verified by comparison with four algorithms each. As verification, mean absolute error was adapted. The presented methods are based on ensemble models using bagging, boosting, and machine learning. The error rate was calculated by assigning the equal weight value and different weight value to each unit model in ensemble. The ensemble model using machine learning showed 61.7% improvement compared to the average of four machine learning technique.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.3
• /
• pp.361-370
• /
• 2017

In this paper, correlation analysis was performed between questionnaire and machine learning based aggressive tendency measurements. this study is part of a aggressive driver detection using machine learning and questionnaire. To collect two types tendency from questionnaire and measurements system, we constructed experiments environments and acquired the data from 30 drivers. In experiment, the machine learning based aggressive tendency measurements system was designed using a driver behavior detection model. And the model was constructed using accelerate and brake position data and hidden markov model method through supervised learning. We performed a correlation analysis between two types tendency using Pearson method. The result was represented to high correlation. The results will be utilize for fusing questionnaire and machine learning. Furthermore, It is verified that the machine learning based aggressive tendency is unique to each driver. The aggressive tendency of driver will be utilized as measurements for advanced driver assistance system such as attention assist, driver identification and anti-theft system.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.6
• /
• pp.827-837
• /
• 2017

Since the area of smart home has been attracting attention, researches have been conducted to control syntagmatically various electronic products with a single remote controller. Previous researches have developed a dedicated controller or an application that acts as a remote controller and controls electronic products by configuring control screen for each product. However, these approaches are not suitable for controlling various electronic products that should be controlled by configuring separate control screens for each product. In this paper, we propose a web-based remote control platform. We define universal user interfaces applicable to various devices by categorizing user interactions of electronic goods and implement them as APIs. By applying the APIs to IPTV and car navigation devices we show that it is possible to control them through only a web browser. We also propose a method to group multiple control requests in order to efficiently handle consecutive control requests and show the improved response time and data usage.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.7 no.10
• /
• pp.935-943
• /
• 2017

CNN (Convolution neural network), which is used for image classification and speech recognition among neural networks learning based on positive data, has been continuously developed to have a high performance structure to date. There are many difficulties to utilize in an embedded system with limited resources. Therefore, we use GPU (General-Purpose Computing on Graphics Processing Units), which is used for general-purpose operation of GPU to solve the problem because we use pre-learned weights but there are still limitations. Since CNN performs simple and iterative operations, the computation speed varies greatly depending on the thread allocation and utilization method in the Single Instruction Multiple Thread (SIMT) based GPGPU. To solve this problem, there is a thread that needs to be relaxed when performing Convolution and Pooling operations with threads. The remaining threads have increased the operation speed by using the method used in the following feature maps and kernel calculations.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.20 no.1
• /
• pp.39-57
• /
• 2009

The two urban-area public libraries and two rural ones that are located in the Gyungsangnam-do province were selected for this paper, and the circulation records in 2007 were collected, and both MS-excel and SPSS were used for their analysis. The collected data were categorized into their collection type, circulation frequencies, and subjects. Also the four libraries were compared and analyzed again for the purpose of comparing the characteristics of the public libraries in urban and rural areas. The number of circulated books, lent number, use factor, and the number of publication lapse year were extracted and analyzed using various types of statistical methods such as correlation coefficient and nonparametric chi-square analysis, etc. as well as descriptive ones.

• Journal of Korea Entertainment Industry Association
• /
• v.13 no.8
• /
• pp.563-572
• /
• 2019

This study was conducted to investigate the effects of school stress and resilience on school children's quality of life. The participants for this study were 266 children of 4, 5, and 6 grades at two elementary schools in G Metropolitan City. Data were collected from September to October, 2019. The participants were assessed for academic stress, resilience, quality of life, and analyzed using descriptive statistics, t-tests, ANOVA, Pearson correlation coefficient, and multiple regression analysis. As a result, factors affecting the quality of life of school-age children were academic stress (β= -.29, p<.001), extracurricular academic stress (β= -.19, p= .004), and resilience (β= .19, p<.001), this variable explained 31.2% of the quality of life of school-age children. The lower the academic stress and the higher the resilience, the higher the quality of life. Therefore, in order to improve the quality of life of school-age children, providers need to develop intervention programs that take these factors into account.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.108-108
• /
• 2022

기후변화는 전 지구 수문순환과 수자원 분포에 커다란 영향을 준다. 하지만 지금까지 관측되어온 지구상의 건조도 변화에 있어서 기후의 자연변동성의 영향과 인간활동에 의한 온난화의 영향을 명시적으로 밝힌 연구는 존재하지 않는다. 본 연구에서는 데이터 구동형 모델과 물리 모델을 이용해 관측 기반의 전구 수자원 분포를 1902년부터 2014년까지 재구축함으로써 지구의 평균온도가 약 1도 상승해온 지난 세기에 걸쳐 건기의 수자원 분포가 어떻게 변해왔는지 보인다. 재구축된 전구 변화 패턴은 인간활동에 의한 온실가스 증가등을 고려한 기후 모델 시뮬레이션과 흡사함을 알 수 있었으며 기후의 자연변동성만을 고려한 기후 모델 시뮬레이션에서는 발견되지 않았다. 주로 북아시아, 북미, 유럽 등 중위도 온대지방에서 더욱더 건조한 건기가 뚜렷하게 나타났으며 이는 강수량의 감소보다는 증발산의 증가에 기인하는 것으로 나타난다. 이와 같은 건조도의 변화는 미래 있어서 또한 인류에 대한 커다란 위협으로 자리한다. 미래 기후에서의 가뭄의 변화에 대해 다양한 연구들이 존재하지만 대부분 높은 수준의 온난화 (예를들어 RCP-SSP 585)에서의 영향에 국한된다. 다시 말해 인류가 21세기 중반에 달성을 목표로 하는 탄소중립이 가뭄의 측면에서 어떤 영향을 주게 될지에 대한 연구는 아직 충분하지 않다고 할 수 있다. 본 연구에서는 약한 혹은 중간 수준의 기후변화 시나리오를 이용해 파리협약에서 목표로 하는 1.5℃와 2℃ 상승에 따라 전 지구의 건조도 분포가 어떻게 변하고 그 변화에 있어서 어떠한 수문기후학적 메커니즘이 작용하는지 밝힌다. 지중해 연안 지역에서는 건조도의 가속이 +1.5℃와 +2℃사이에 존재하였으나 동아시아에서는 +1.5℃와 +2℃ 모두에서 습윤해짐을 알 수 있었으며 이러한 지역적 불균일성은 기후변화 대응 노력에 있어서 과거 온실가스 배출에 대한 책임뿐만 아니라 다양한 부문에 걸친 미래의 잠재 적응 노력 또한 고려해야만 함을 시사한다. 본 연구는 제6차 Coupled Model Intercomparison Project의 Land Surface, Snow, Soil-moisture Model Intercomparison Project (CMIP6/LS3MIP)와 Half a degree Additional warming, Prognosis and Projected Impacts (HAPPI)의 다중 모델 앙상블 시뮬레이션 결과를 이용했다.

• Smart Media Journal
• /
• v.11 no.7
• /
• pp.61-69
• /
• 2022

In the learned machine learning, the performance of machine learning degrades at the same time as drift occurs in terms of learning models and learning data over time. As a solution to this problem, I would like to propose the concept and evaluation method of ML drift to determine the re-learning period of machine learning. An XAI test and an XAI test of an apple image were performed according to strawberry and clarity. In the case of strawberries, the change in the XAI analysis of ML models according to the clarity value was insignificant, and in the case of XAI of apple image, apples normally classified objects and heat map areas, but in the case of apple flowers and buds, the results were insignificant compared to strawberries and apples. This is expected to be caused by the lack of learning images of apple flowers and buds, and more apple flowers and buds will be studied and tested in the future.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.4
• /
• pp.1-14
• /
• 2023

As markets and industries continue to evolve rapidly, technology opportunity discovery (TOD) has become critical to a firm's survival. From a common consensus that TOD based on a firm's capabilities is a valuable method for small and medium-sized enterprises (SMEs) and reduces the risk of failure in technology development, studies for TOD based on a firm's capabilities have been actively conducted. However, previous studies mainly focused on a firm's technological capabilities and rarely on business capabilities. Since discovered technologies can create market value when utilized in a firm's business, a firm's current business capabilities should be considered in discovering technology opportunities. In this context, this study proposes a TOD method that considers both a firm's business and technological capabilities. To this end, this study uses patent data, which represents the firm's technological capabilities, and trademark data, which represents the firm's business capabilities. The proposed method comprises four steps: 1) Constructing firm technology and business capability matrices using patent classification codes and trademark similarity group codes; 2) Transforming the capability matrices to preference matrices using the fuzzy function; 3) Identifying a target firm's candidate technology opportunities using the collaborative filtering algorithm; 4) Recommending technology opportunities using a portfolio map constructed based on technology similarity and applicability indices. A case study is conducted on a security firm to determine the validity of the proposed method. The proposed method can assist SMEs that face resource constraints in identifying technology opportunities. Further, it can be used by firms that do not possess patents since the proposed method uncovers technology opportunities based on business capabilities.