• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.2
• /
• pp.202-207
• /
• 2010

In this paper, we present an automatic sentiment classification method for on-line movie reviews that do not contain explicit sentiment rating scores. For the sentiment polarity classification, positive or negative, we use a Support Vector Machine classifier based on syllable kernel that is an extended model of string kernel. We give some experimental results which show that proposed syllable kernel model can be effectively used in sentiment classification tasks for on-line movie reviews that usually contain a lot of grammatical errors such as spacing or spelling errors.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.4 no.1
• /
• pp.49-55
• /
• 2005

This paper presents experiments on the evaluation of characteristics of ER fluids used for vibration control of application in intelligence type process control systems. Dynamic characteristics of the actuator(beam) embedded with the ER fluid can be controlled by changing the strength of the electric field applied on the ER fluids, thus provides a mean to avoid the resonance. In case electric field is supplied to the smart structure with ER fluids, vibration energy is dissipated more than the beam without electric field, because particles in ER fluid form a chain structure in the presence of electric field. The damping and stiffness of the beam with ER fluid are increased when the applied electric field increases. The characteristics of damping and stiffness of the ER fluid with various electric field strength were investigated by conducting a vibration test of the beam with ER fluid. If it applies characteristics of the ER fluids, it will be able to apply in the PLC control system for the vibration which occurs from process system.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.3
• /
• pp.291-296
• /
• 2007

The ontology has been used widely in recent years with its aim to accumulate knowledge that machine can comprehend. We believe that machine can manage and analyze information on its own using the ontology. In this paper, we propose an algorithm that allows us to extract properties of ontology instances from structured information already existing in web documents. In particular, by stratification of the domain knowledge that is composed of property information, we were able to make the algorithm better and improve the quality of extraction results. In our experiments with 20 thousands targeted documents, we were able to extract property information with 83% confidence.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.20 no.4
• /
• pp.1133-1145
• /
• 1996

A method is proposed to predict the deformed shape of the structure subjected to the unknown external loads using the signal from the piezoceramic sensors. Such a shape estimation is based on the linear relationship between the deformation of structure and the signal from sensor, which is calculated using finite element method. The deformed shape is, then calculated using the linear matrix and the signals from the piezoceramic sensors attached to the structures. For the purpose, a structural analysis program is developed using a multi-layerd finite element of 8 nodes with 3 displacement and one voltage degrees of freedom at each node. The multiple layers with the different material properties can be layered within the element. The incompatible mode with the element is found to be crucial to catch the bending behavior accurately. The accuracy of the program is, then, verified by being compared with the experimental results performed by Crawley. The proposed shape estimation method is also verified for the different loads and sensor size. It is shown that the results of shape estimation method using the linear matrix well predicts the deflections compared with those of finite element method.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.06a
• /
• pp.383-389
• /
• 2001

된 연구에서 대상으로 하고 있는 LG 정유 광양항 제품부두는 7 선석(Berth)에 재화중량(DWT) 300톤에서 48000 톤의 선박까지 다양한 선박이 이용하고 있으며, 해상의 기상상태에 따른 선박 입출향 통제 지침 설정이 어렵고, 현재 사용하고 있는 지침의 근거가 명확하지 않아 현재의 부두 운영이 비효율적이거나 안전성이 결여되어 있다고 할 수 있다. 따라서 이를 개선하기 위한 합리적인 부두운영 제한조건 개발이 절실히 요구되었다. 본 논문에서는 대상 부두의 특성, 대상 선박의 특성, 하중상태, 선박 운항자의 특성 등을 고려하여 해상/기상 상황(바람, 조류 및 파랑)에 따른 부두 입출항 가능 여부를 정량적으로 판단하고, 안전성 향상 방안을 제시할 수 있는 의사결정 시스템을 개발하고 5번, 7번 선석을 대상으로 이를 검증하였다. 여기서는 입출항 여부를 정량적으로 판단하여 결과를 제시하기 위해서 유전적 프로그래밍(Genetic Programming)을 이용한 기계학습 방법을 이용하였으며, GP의 방대한 계산량을 줄이기 위한 가중 선형 연상 기억(Weighted Linear Associative Memory: WLAM) 방법의 도입 및 전역 최적점을 쉽게 찾기 위한 Group of Additive Genetic Programming Trees(GAGPT)를 도입함으로써 학습 성능을 개선하였다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.10a
• /
• pp.954-957
• /
• 2005

In order to implement Artificial Intelligence, various technologies have been widely used. Artificial Intelligence is applied for many industrial product and machine tools are the center of manufacturing devices in intelligent manufacturing system. The purpose of this paper is to present the construction of Rule Base for knowledge structure that is applicable to machine tools. This system is that decision whether to act in accordance with machine status is support system. It constructs Rule Base of knowledge used of machine toots. The constructed Rule Base facilitates the effective operation and control of machine tools and will provide a systematic way to integrate the expert's knowledge that will apply Intelligent Machine Tools.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.3
• /
• pp.341-346
• /
• 2011

A vacant technology forecasting is an important issue in management of technology. The forecast of vacant technology leads to the growth of nation and company. So, we need the results of technology developments until now to predict the vacant technology. Patent is an objective thing of the results in research and development of technology. We study a predictive method for forecasting the vacant technology quantitatively using patent data in this paper. We propose an ensemble model that is to vote some clustering criteria because we can't guarantee a model is optimal. Therefore, an objective and accurate forecasting model of vacant technology is researched in our paper. This model combines statistical analysis methods with machine learning algorithms. To verify our performance evaluation objectively, we make experiments using patent documents of diverse technology fields.

• Particle and aerosol research
• /
• v.18 no.4
• /
• pp.97-107
• /
• 2022

The geometry of popular wave plate type mist eliminator for the wet flue gas desulfurization process was improved, fabricated, and experimentally evaluated. A Mist eliminator is a type of inertial particle collector which collection efficiency is proportional to the velocity of the gas phase. However, as the amount of re-entrainment is also proportional to the gas phase velocity, there is a limitation for the gas phase flow rate. Re-entrainment is one of the most important issues in a mist eliminator and is likely to occur as the input of the liquid phase and flow rate of the gas phase increase. In order to resolve this problem, the projection angle of the improved mist eliminator is set to 30° from the conventional one while maintaining the cross-section. With low flow rate conditions, the modified mist eliminator showed a similar pressure drop and overall collection efficiency. However, with conditions in which re-entrainment is obviously occurring, the modified mist eliminator showed better performance in draining droplets than the conventional one. As a result, the modified mist eliminator showed higher overall collection efficiency.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.245-248
• /
• 2022

최근 자연어 처리 분야에서 문맥적 의미를 반영하기 위한 대조학습 (contrastive learning) 에 대한 연구가 활발히 이뤄지고 있다. 이 때 대조학습을 위한 양질의 학습 (training) 데이터와 검증 (validation) 데이터를 이용하는 것이 중요하다. 그러나 한국어의 경우 대다수의 데이터셋이 영어로 된 데이터를 한국어로 기계 번역하여 검토 후 제공되는 데이터셋 밖에 존재하지 않는다. 이는 기계번역의 성능에 의존하는 단점을 갖고 있다. 본 논문에서는 한국어 리뷰 데이터로 임베딩의 의미 반영 정도를 측정할 수 있는 간단한 검증 데이터셋 구축 방법을 제안하고, 이를 활용한 데이터셋인 SimKoR (Similarity Korean Review dataset) 을 제안한다. 제안하는 검증 데이터셋을 이용해서 대조학습을 수행하고 효과성을 보인다.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.5
• /
• pp.217-220
• /
• 2024

Predicting earthquake occurrences accurately is challenging, and preparing all buildings with seismic design for such random events is a difficult task. Analyzing building features to predict potential damage and reinforcing vulnerabilities based on this analysis can minimize damages even in buildings without seismic design. Therefore, research analyzing the efficiency of building damage prediction models is essential. In this paper, we compare the accuracy of earthquake damage prediction models using machine learning classification algorithms, including Random Forest, Extreme Gradient Boosting, LightGBM, and CatBoost, utilizing data from buildings damaged during the 2015 Nepal earthquake.