• Journal of Digital Convergence
• /
• v.17 no.3
• /
• pp.107-117
• /
• 2019

In the era of SNS, many people relies on it to express their emotions about various kinds of products and services. Therefore, for the companies eagerly seeking to investigate how their products and services are perceived in the market, emotion mining tasks using dataset from SNSs become important much more than ever. Basically, emotion mining is a branch of sentiment analysis which is based on BOW (bag-of-words) and TF-IDF. However, there are few studies on the emotion mining which adopt feature selection (FS) methods to look for optimal set of features ensuring better results. In this sense, this study aims to propose FS methods to conduct emotion mining tasks more effectively with better outcomes. This study uses Twitter and SemEval2007 dataset for the sake of emotion mining experiments. We applied three FS methods such as CFS (Correlation based FS), IG (Information Gain), and ReliefF. Emotion mining results were obtained from applying the selected features to nine classifiers. When applying DT (decision tree) to Tweet dataset, accuracy increases with CFS, IG, and ReliefF methods. When applying LR (logistic regression) to SemEval2007 dataset, accuracy increases with ReliefF method.

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.427-438
• /
• 2021

Two-stage sampling allows us to estimate population characteristics by both unit and cluster level together. Given a complex auxiliary information, integrated calibration weighting would better reflect the level-wise characteristics as well as multivariate characteristics between levels. This paper explored the integrated calibration weighting methods by Estevao and Särndal (2006) and Kim (2019) through a simulation study, where the efficiency of those weighting methods was compared using an artificial population data. Two weighting methods among others are shown efficient: single step calibration at the unit level with stacked individualized auxiliary information and iterative integrated calibration at each level. Under both methods, cluster calibrated weights are defined as the average of the calibrated weights of the unit(s) within cluster. Both were very good in terms of the goodness-of-fit of estimating the population totals of mutual auxiliary information between clusters and units, and showed small relative bias and relative mean square root errors for estimating the population totals of survey variables that are not included in calibration adjustments.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06a
• /
• pp.155-157
• /
• 2012

최근 클라우드 컴퓨팅의 핵심 기술로서 가상화 기술이 대두되고 있다. 가상화는 컴퓨터 자원의 물리적인 특징을 추상화하여 사용자에게는 논리적 자원을 제공하고 이를 통해 기술적 및 관리적 이점을 제공한다. Xen 가상화 환경에서 가상머신에 디스크 공간을 할당하는 한 가지 방식으로 하나의 하드디스크 파티션을 가상머신에 할당하는 방식이 있다. 위와 같은 환경에서는 같은 하드디스크에 존재하는 가상머신들임에도 불구하고 데이터에 접근하는 물리적 특성으로 인해 읽기/쓰기 성능 불균형이 발생하고, 가상머신의 수가 증가함에 따라 파티션 간의 최단 탐색거리 증가로 인해 동시적 읽기/쓰기의 성능이 저하되는 문제점이 발생한다. 본 논문은 이러한 문제점을 해결하기 위해 새로운 파티션 구조인 Onion 파티션 구조를 제안한다. Xen버전 4.1의 환경에서 Domain-0 커널버전 2.6.32를 수정하여 구현 및 실험을 하였다. 실험 결과, 각 가상머신의 독립적 읽기/쓰기 성능 편차가 기존대비 8%, 10%수준으로 개선되었으며, 가상머신들의 동시적 읽기/쓰기 성능은 기존대비 156.76%, 119.01%의 성능 향상을 보였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06a
• /
• pp.98-100
• /
• 2012

모바일 웜의 전파 속도는 시간이 지날수록 급격히 빨라진다. 급속도로 전파하는 웜을 억제하기 위한 연구가 계속 진행되어 왔다. 기존 연구 중 하나로 소셜 네트워크를 그래프(Graph)로 표현하고 분할한 뒤 각 파티션의 대표노드(Key Node)를 선택해 웜 패치를 전파하는 기법이 있다. 이 기법은 패치 전파 속도를 웜 전파 속도보다 빠르게 해서 웜을 억제 시킨다. 하지만 이 기법에서는 대표노드에 가중치(Weight)가 낮은 간선(Edge)이 연결 될 수 있거나 대표노드가 없는 파티션이 생길 수 있다. 잘못된 대표노드 선정은 패치 전파속도를 저하시키고 웜 억제를 지연시킬 수 있다. 본 논문에서는 페이지랭크를 이용해서 대표노드를 선정하는 개선된 소셜 네트워크 기반의 웜 패칭 기법을 제안한다. 제안 기법은 웜 패칭 전략에 사용할 수 있도록 수정한 페이지랭크를 각 파티션에 적용해서 대표노드를 선정한다. 파티션마다 전파 능력이 가장 높은 노드가 대표노드로 선정되기 때문에 웜 패치 전파 속도를 향상시킬 수 있다.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.4 s.175
• /
• pp.955-962
• /
• 2000

Scanning Probe Microscope (SPM) such as Scanning Tunneling Microscope (STM) and Atomic Force Microscope (AFM) was shown to be the powerful tool for nano-scale characterization of a fracture surface . AFM was used to study cross sectional profiles and dimensions of fatigue striations in 2017-T351 aluminum alloy. Their widths (SW) and heights (SH) were measured from the cross sectional profiles of three-dimension AFM images. The following results that will be helpful to understand the fatigue crack growth mechanism were obtained. (1) Coincidence of the crack growth rate with the striation width was found down to the growth rate of 10-5 mm/cycle. (2) The relation of SH=0.085(SW)1.2 was obtained. (3) The ratio of the striation height to its width SH/SW did not depend on the stress intensity factor range K and the stress ratio R. (4) Not only the SW but also the SH changed linearly with the crack tip opening displacement (CTOD) when plotted in log-log scale. From these results, the applicability of the AFM to nano-fractography is discussed.

• The Transactions of the Korea Information Processing Society
• /
• v.3 no.1
• /
• pp.87-94
• /
• 1996

In order to design an extended relational database management system supporting multilevel security, the standard relational data model is extended and new relational integrity constrains are proposed for the model. The extended relational model and proposed multilevel integrity constraniants maintain database in consistent state and produce a basis that can eliminat eambiguity of entity and relation ship representations bypoly instantiation. The proposed up dates emantics canincreases the efficiency of up date operations by supporting multilevel entry and up dates. The semantics also provides a basis for the implementation of decomposition of extended relations.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.5 no.4
• /
• pp.263-270
• /
• 1993

A study on the variation of radiation stress and mean water level is carried out for the shoaling and breaking waves on a plane beach. In general, the radiation stresses computed based on the linear wave theory are overestimated. which results in the discrepancy between the computed results and laboratory data of mean water level in the surf zone. In this paper, by modifying the Svendsen's approach (1984), radiation stress is expressed in terms of water depth. The computed results are compared with the results calculated by a linear wave theory and Sawaragi's approach (1984) based on the spectrum of breaking wave components, and published laboratory data. The computed results of the modifed Svendsen's approach are favourably compared with the laboratory data.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.22 no.3
• /
• pp.223-229
• /
• 2009

Most of structural analyses are concerned with the deformation and stress in a body subjected to external loads. In many fields, however, the interpretation of inverse problems is needed to determine surface tractions or internal stresses from measured displacements. In this study, the inverse processes by using the boundary element method are formulated for the evaluation of boundary tractions from displacements measured on a remote surface. Small errors in measured displacements often result in a substantial loss of accuracy of an inverse system. Numerical results show that the error in reconstructed tractions by using the inverse boundary element methods is sensitive to measurement location and noise.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.8
• /
• pp.1336-1345
• /
• 2002

Materialized views in data warehouse play important roles in rapidly answering to users's requests for information Processing. More views in data warehouse, can respond to the users more rapidly because the user's requests may be processed by using only the materialized views with higher probabilities rather than accessing base relations. The limited duration, during which the materialized views are updated due to base relations's changes, limits the number of materialized views in data warehouse. In this paper, we propose efficient policy for updating the materialized views, which can save the update duration of views although MIN/MAX values frequently change in base relation. The policy updates the materialized views by distinguishing whether MIN/MAX values's changes in the base relation are insert value or delete value. Then, the number of accesses to the base relation is descreased when updating the MIN/MAX values in the materialized views.

• Magazine of the Korean Society of Agricultural Engineers
• /
• v.36 no.2
• /
• pp.88-95
• /
• 1994

This paper is to evaluate the applicability of neural networks to the estimation of evapotranspiration. Two neural networks were developed to forecast daily evapotranspiration of the rice crop with back-propagation and counter-propagation algorithm. The neural network trained by back-propagation algorithm with delta learning rule is a three-layer network with input, hidden, and output layers. The other network with counter-propagation algorithm is a four-layer network with input, normalizing, competitive, and output layers. Training neural networks was conducted using daily actual evapotranspiration of rice crop and daily climatic data such as mean temperature, sunshine hours, solar radiation, relative humidity, and pan evaporation. During the training, neural network parameters were calibrated. The trained networks were applied to a set of field data not used in the training. The created response of the back-propagation network was in good agreement with desired values and showed better performances than the counter-propagation network did. Evaluating the neural network performance indicates that the back-propagation neural network may be applied to the estimation of evapotranspiration of the rice crop. This study does not provide with a conclusive statement as to the ability of a neural network to evapotranspiration estimating. More detailed study is required for better understanding and evaluating the behavior of neural networks.