• ETRI Journal
• /
• v.43 no.2
• /
• pp.299-312
• /
• 2021

It is necessary to achieve high performance in the task of zero anaphora resolution (ZAR) for completely understanding the texts in Korean, Japanese, Chinese, and various other languages. Deep-learning-based models are being employed for building ZAR systems, owing to the success of deep learning in the recent years. However, the objective of building a high-quality ZAR system is far from being achieved even using these models. To enhance the current ZAR techniques, we fine-tuned a pretrained bidirectional encoder representations from transformers (BERT). Notably, BERT is a general language representation model that enables systems to utilize deep bidirectional contextual information in a natural language text. It extensively exploits the attention mechanism based upon the sequence-transduction model Transformer. In our model, classification is simultaneously performed for all the words in the input word sequence to decide whether each word can be an antecedent. We seek end-to-end learning by disallowing any use of hand-crafted or dependency-parsing features. Experimental results show that compared with other models, our approach can significantly improve the performance of ZAR.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.9
• /
• pp.2334-2347
• /
• 2023

Metaverse services generate text data, data of ubiquitous computing, in real-time to analyze user emotions. Analysis of user emotions is an important task in metaverse services. This study aims to classify user sentiments using deep learning and pre-trained language models based on the transformer structure. Previous studies collected data from a single platform, whereas the current study incorporated the review data as "Metaverse" keyword from the YouTube and Google Play Store platforms for general utilization. As a result, the Bidirectional Encoder Representations from Transformers (BERT) and Robustly optimized BERT approach (RoBERTa) models using the soft voting mechanism achieved a highest accuracy of 88.57%. In addition, the area under the curve (AUC) score of the ensemble model comprising RoBERTa, BERT, and A Lite BERT (ALBERT) was 0.9458. The results demonstrate that the ensemble combined with the RoBERTa model exhibits good performance. Therefore, the RoBERTa model can be applied on platforms that provide metaverse services. The findings contribute to the advancement of natural language processing techniques in metaverse services, which are increasingly important in digital platforms and virtual environments. Overall, this study provides empirical evidence that sentiment analysis using deep learning and pre-trained language models is a promising approach to improving user experiences in metaverse services.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.56 no.3
• /
• pp.472-480
• /
• 2007

This paper analyzes overvoltage on testing line for various parameter effect examination. Model systems consist of overhead line and underground cable. The model considered actual characteristic data of distribution lines. and will be constructed at testing yard. The simulations were performed under various line configuration such as cable kinds, cable length, lightning wave, lightning wave time, transformer and branch circuits. The simulation models are established by EMTP/ATPDraw and Line Constants are calculated by ATP_LCC. When lightning surge strikes on conductor of overhead line, EMTP/ATPDraw calculates overvoltage in many cases. Simulation results will be compared with real testing results at testing yard. The compared results will be used to establish protection methods in actual underground distribution systems.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.519-520
• /
• 2011

In Gas Insulated Substation(GIS), Very Fast Transient Overvoltages(VFTOs) are generated by the switching operation of disconnector or circuit breaker and it is propagated to each component of GIS. Each component of GIS is modeled by lumped line model and distributed line model based on equivalent circuits. The various cases which can be operated and the various closing point-on-wave are considered. In this paper, magnitude and rate-of-rise of VFTO are simulated and analyzed by comparing the simulation results. Measurements are performed at transformer and circuit breaker.

• KIEE International Transactions on Power Engineering
• /
• v.5A no.4
• /
• pp.366-370
• /
• 2005

To meet increasing power demand, 500 kV power systems are under consideration in the regions of some Middle Asian countries. As the power system voltage becomes higher, the cost for the power system insulation increases significantly. 500 kV transmission systems will become the basis of a region's power system and they require much higher system reliability. Consequently, by the methods of limiting overvoltages effectively, a reasonable insulation design and coordination must be accomplished. In particular, the Substations must be constructed to be of outdoor type. In order to determine the various factors for the insulation design, the EMTP (Electro-magnetic transient program) is used for the magnification of transient phenomena of the 500 kV systems in the planned network. In this paper, we will explain the calculation results of lightning overvoltages by the EMTP for lightning protection design for the 500 kV substations. To obtain reliable results, the multi-story tower model and EMTP/TACS model are introduced for the simulation of dynamic arc characteristics.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.41 no.4
• /
• pp.331-338
• /
• 1992

The application of Connectionist expert system using neural network to fault diagnosis of power system is presented and compared with rule-based expert system. Also, the merits of Connectionist model using neural network is presented. In this paper, the neural network for fault diagnosis is hierarchically composed by 3 neural network classes. The whole power system is divided into subsystems, the neural networks (Class II) which take charge of each subsystem and the neural network (Class III) which connects subsystems are composed. Every section of power system is classified into one of the typical sections which can be applied with same diagnosis rules, as line-section, bus-section, transformer-section. For each typical section, only one neural network (Class I) is composed. As the proposed model has hierarchical structure, the great reduction of learning structure is achieved. With parallel distributed processing, we show the possibility of on-line fault diagnosis.

• Proceedings of the KIEE Conference
• /
• 2002.07a
• /
• pp.541-543
• /
• 2002

This paper describes the analysis resulates on the P.T ferroresonance at 154kV GIS(Gas Insulated Substation) by EMTP(Electromagnetic transient program). We had simulated the PT ferroresonance between a potential transformer(PT) and open circuit braker's grading capacitance. The ferroresonance leads to very large power frequency overvoltage on PT bus and subsequent insulation failure. The damping circuit connected on the secondary winding were considered in the model. The actual countermeasure include a saturable inductor and a resistor series mounted, but the 2 ohms of damping resistance was used in the computational model.

• Proceedings of the KIEE Conference
• /
• 2000.11a
• /
• pp.199-201
• /
• 2000

This paper describes the theory and experimental result of scaled model of STATCOM. The STATCOM, a solid-state voltage source inverter coupled with a transformer, is tied to a transmission line. A STATCOM inject an almost sinusoidal current of variables magnitude, at the pont of connection. This injected current is almost in quadrature with the line voltage, thereby emulating an inductive or capacitive reactance at the point of connection with the transmission line.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2000.05b
• /
• pp.216-221
• /
• 2000

This paper describes the analysis results on the PT ferroresonance at 154 kV GIS (gas Insulated substation by EMTP(Electro- magnetic Transient Program). We had simulated the PT ferroresonance between a potential transformer(PT) and an open circuit breaker's grading capacitance. The ferroresonance leads to very large power frequency overvoltages on PT bus and subsequent insulation failure. The large power can be supplied to the PT by the high voltage on the opened circuit due to the grading capacitance and equivalent capacitance of the buses during the ferroresonance. The damping circuit connected on the secondary winding were considered in the model. The actual countermeasure include a saturable inductor and a resistor series-mounted, but the 2 ohms of damping resistance was used in the computationl model.

• Proceedings of the KSR Conference
• /
• 2005.05a
• /
• pp.794-799
• /
• 2005

In this paper, the inductive power collector using electromagnetic induction for the PRT(Personal Rapid Transit) system is suggested and some ideas for power collector design to improve the power transfer performance are presented. The proposed the inductive power collector is used for the PRT system, which has a large air-gap and demands a large electrical power capability. But, low output power is generated due to a loosely coupled characteristic of the large air-gap. Therefore, double layer construction of secondary winding. which was divided in half to increase both output current and output voltage was suggested. Also, model of power collector and parallel winding structure are presented, in addition, the performance of inductive power collector to alignment condition between the primary power line and the inductive power transformer was verified by computer simulation of 2kW model.