• Journal of Information Processing Systems
• /
• v.18 no.4
• /
• pp.535-548
• /
• 2022

Using a rich resource language to classify sentiments in a language with few resources is a popular subject of research in natural language processing. Burmese is a low-resource language. In light of the scarcity of labeled training data for sentiment classification in Burmese, in this study, we propose a method of transfer learning for sentiment analysis of a language that uses the feature transfer technique on sentiments in English. This method generates a cross-language word-embedding representation of Burmese vocabulary to map Burmese text to the semantic space of English text. A model to classify sentiments in English is then pre-trained using a convolutional neural network and an attention mechanism, where the network shares the model for sentiment analysis of English. The parameters of the network layer are used to learn the cross-language features of the sentiments, which are then transferred to the model to classify sentiments in Burmese. Finally, the model was tuned using the labeled Burmese data. The results of the experiments show that the proposed method can significantly improve the classification of sentiments in Burmese compared to a model trained using only a Burmese corpus.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.1996-2015
• /
• 2023

High-definition (HD) maps can provide precise road information that enables an autonomous driving system to effectively navigate a vehicle. Recent research has focused on leveraging semantic segmentation to achieve automatic annotation of HD maps. However, the existing methods suffer from low recognition accuracy in automatic driving scenarios, leading to inefficient annotation processes. In this paper, we propose a novel semantic segmentation method for automatic HD map annotation. Our approach introduces a new encoder, known as the convolutional transformer hybrid encoder, to enhance the model's feature extraction capabilities. Additionally, we propose a multi-level fusion module that enables the model to aggregate different levels of detail and semantic information. Furthermore, we present a novel decoupled boundary joint decoder to improve the model's ability to handle the boundary between categories. To evaluate our method, we conducted experiments using the Bird's Eye View point cloud images dataset and Cityscapes dataset. Comparative analysis against stateof-the-art methods demonstrates that our model achieves the highest performance. Specifically, our model achieves an mIoU of 56.26%, surpassing the results of SegFormer with an mIoU of 1.47%. This innovative promises to significantly enhance the efficiency of HD map automatic annotation.

• Journal of the Korean Geotechnical Society
• /
• v.20 no.2
• /
• pp.125-130
• /
• 2004

In this paper, the method of parameter estimation of a mathematical constitutive model blown as the smooth elasto-plastic cap model is studied. To predict the response of the real soil using this model, the eight parameters describing the constitutive equations have to be determined. First, experimental data are obtained from simple laboratory experiments such as one dimensional confined compression test in a consolidometer and drained triaxial compression test with the Ottawa sand f3r the reference value. Then, the numerical experiments are performed in the cap model with initial guessed parameters. The optimization method is utilized to fit the model response to experimental data by minimizing the error between the laboratory and numerical responses. Special attention is given to the parameter estimation procedure of numerical triaxial test due to the difficulty of the lateral strain measurements.

• ETRI Journal
• /
• v.32 no.4
• /
• pp.634-637
• /
• 2010

This letter presents a model for a dynamic collaboration (DC) platform among cloud providers (CPs) that prevents adverse business impacts, cloud vendor lock-in and violation of service level agreements with consumers, and also offers collaborative cloud services to consumers. We consider two major challenges. The first challenge is to find an appropriate market model in order to enable the DC platform. The second is to select suitable collaborative partners to provide services. We propose a novel combinatorial auction-based cloud market model that enables a DC platform among CPs. We also propose a new promising multi-objective optimization model to quantitatively evaluate the partners. Simulation experiments were conducted to verify both of the proposed models.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2001.03a
• /
• pp.591-596
• /
• 2001

The purpose of this paper Is to present and discuss some of evaporation and desiccation observed in laboratory experiments under controlled conditions, and is to improve PTM(Progressive Trenching Method) operating technique. PTM is the technically feasible and economically justifiable dewatering and desiccation technique for dredged material containment areas. A series of laboratory experiments with large model test were carried out to get evaporation rate and strength increase. Surface desiccation of dredged material is basically changed by evaporation characteristics which is controlled by weather and trench type, etc. This study shows that trench depth and rain fall are important factors in desiccation of dredged soft clay.

• Proceedings of the Korea Committee for Ocean Resources and Engineering Conference
• /
• 2004.11a
• /
• pp.212-219
• /
• 2004

The effect of frequency and amplitude of the OWC (Oscillating Water Column) motion on the nonlinear reaction forces in an air duct arc studied experimentally. Experimental OWC model is idealized as a simple circular cylinder with an orifice type air duct located at the middle of the top rid. Reaction forces due to forced heave oscillation are measured and analyzed. By subtracting the effect of inertia forces and restoring forces, pneumatic damping force and added spring force are deduced. The effects of the frequency and amplitude of the heave motion are discussed. Also, the effects of solidity of the duct on the reaction forces are discussed.

• Water Engineering Research
• /
• v.2 no.1
• /
• pp.21-31
• /
• 2001

In this study, laboratory experiments were performed to investigate the flow characteristics of a two-dimensional neutrally buoyant jet in the inlet region of a rectangular laboratory settling tank. Velocity measurements were made with a three-component ADV. Two types of baffles were installed in front of two-dimensional slot; a one-sided and a two-sided baffle. The flow fields from a plane jet impinging on these two types of baffles and a plane jet without a baffle showed quite different characteristics. To concentrate on investigating these flow characteristics, the effects of density currents due to temperature difference or the presence of sediments were not studied. Results of the experiments reveal that the use of the two-sided baffle results in the shortest inlet region. Also shown is that, in addition to the types of baffles, the Froude number turns out to be an important factor in the extent of the inlet region.

• Smart Structures and Systems
• /
• v.26 no.1
• /
• pp.63-75
• /
• 2020

Substructure pseudo-dynamic hybrid simulation (SPDHS) combining the advantages of physical experiments and numerical simulation has become an important testing method for evaluating the dynamic responses of structures. Various parameter identification methods have been proposed for online model updating. However, if there is large model gap between the assumed numerical models and the real models, the parameter identification methods will cause large prediction errors. This study presents an ANN (artificial neural network) method based on forgetting factor. During the SPDHS of model updating, a dynamic sample window is formed in each loading step with forgetting factor to keep balance between the new samples and historical ones. The effectiveness and anti-noise ability of this method are evaluated by numerical analysis of a six-story frame structure with BRBs (Buckling Restrained Brace). One BRB is simulated in OpenFresco as the experimental substructure, while the rest is modeled in MATLAB. The results show that ANN is able to present more hysteresis behaviors that do not exist in the initial assumed numerical models. It is demonstrated that the proposed method has good adaptability and prediction accuracy of restoring force even under different loading histories.

• Environmental Engineering Research
• /
• v.14 no.3
• /
• pp.164-169
• /
• 2009

A powder form of aluminum (hydr)oxides is not suitable in wastewater treatment/filtration systems because of low hydraulic conductivity and large sludge production. In this study, aluminum (hydr)oxide-coated sand (AOCS) was used to remove phosphate from aqueous solution. The properties of AOCS were analyzed using a scanning electron microscopy (SEM) combined with an energy dispersive X-ray spectrometer (EDS) and an X-ray diffractometer (XRD). Kinetic batch, equilibrium batch, and closed-loop column experiments were performed to examine the adsorption of phosphate to AOCS. The XRD pattern indicated that the powder form of aluminum (hydr)oxides coated on AOCS was similar to a low crystalline boehmite. Kinetic batch experiments demonstrated that P adsorption to AOCS reached equilibrium after 24 h of reaction time. The kinetic sorption data were described well by the pseudo second-order kinetic sorption model, which determined the amount of P adsorbed at equilibrium ($q_e$ = 0.118 mg/g) and the pseudo second-order velocity constant (k = 0.0036 g/mg/h) at initial P concentration of 25 mg/L. The equilibrium batch data were fitted well to the Freundlich isotherm model, which quantified the distribution coefficient ($K_F$ = 0.083 L/g), and the Freundlich constant (1/n = 0.339). The closed-loop column experiments showed that the phosphate removal percent decreased from 89.1 to 41.9% with increasing initial pH from 4.82 to 9.53. The adsorption capacity determined from the closed-loop experiment was 0.239 mg/g at initial pH 7.0, which is about two times greater than that ($q_e$ = 0.118 mg/g) from the kinetic batch experiment at the same condition.

• Atmosphere
• /
• v.18 no.4
• /
• pp.299-316
• /
• 2008

The intensive upper-air observation network was organized over southwestern region of the Korean Peninsula during the Korea Enhanced Observing Program in 2005 (KEOP-2005). In order to examine the effect of additional upper-air observation on the numerical weather forecasting, three Observing System Experiments (OSEs) using Korea Local Analysis and Prediction System (KLAPS) and Weather Research and Forecasting (WRF) model with KEOP-2005 data are conducted. Cold start case with KEOP-2005 data presents a remarkable predictability difference with only conventional observation data in the downstream and along the Changma front area. The sensitivity of the predictability tends to decrease under the stable atmosphere. Our results indicates that the effect of intensive observation plays a role in the forecasting of the sensitive area in the numerical model, especially under the unstable atmospheric conditions. When the intensive upper-air observation data (KEOP-2005 data) are included in the OSEs, the predictability of precipitation is partially improved. Especially, when KEOP-2005 data are assimilated at 6-hour interval, the predictability on the heavy rainfall showing higher Critical Success Index (CSI) is highly improved. Therefore it is found that KEOP-2005 data play an important role in improving the position and intensity of the simulated precipitation system.