• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.10
• /
• pp.455-460
• /
• 2023

Text augmentation is a methodology that creates new augmented texts by transforming or generating original texts for the purpose of improving the performance of NLP models. However existing text augmentation techniques have limitations such as lack of expressive diversity semantic distortion and limited number of augmented texts. Recently text augmentation using large language models and few-shot learning can overcome these limitations but there is also a risk of noise generation due to incorrect generation. In this paper, we propose a text augmentation method called TAGS that generates multiple candidate texts and selects the appropriate text as the augmented text. TAGS generates various expressions using few-shot learning while effectively selecting suitable data even with a small amount of original text by using contrastive learning and similarity comparison. We applied this method to task-oriented chatbot data and achieved more than sixty times quantitative improvement. We also analyzed the generated texts to confirm that they produced semantically and expressively diverse texts compared to the original texts. Moreover, we trained and evaluated a classification model using the augmented texts and showed that it improved the performance by more than 0.1915, confirming that it helps to improve the actual model performance.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.43 no.6
• /
• pp.775-784
• /
• 2023

In terms of flood management, it is necessary to analyze quantitative rainfall and runoff from a spatial and temporal perspective and to analyze runoff for heavy rainfall events that are concentrated within a short period of time. The simulation and analysis results of rainfall-runoff models vary depending on the type and input data. In particular, rainfall data is an important factor, so calculating areal mean rainfall is very important. In this study, the areal mean rainfall of the Samcheok Osipcheon(Riv.) watersheds located in the mountainous terrain was calculated using the Arithmetic Mean Method, Thiessen's Weighting Method, and the Isohyetal Method, and the rainfall-runoff results were compared by applying the distributional model S-RAT and the lumped model HEC-HMS. The results of the temporal transferability study showed that the combination of the distributional model and the Isohyetal Method had the best statistical performance with MAE of 64.62 m³/s, RMSE of 82.47 m³/s, and R² and NSE of 0.9383 and 0.8547, respectively. It is considered that this study was properly analyzed because the peak flood volume occurrence time of the observed and simulated flows is within 1 hour. Therefore, the results of this study can be used for frequency analysis in the future, which can be used to improve the accuracy of simulating peak flood volume and peak flood occurrence time in mountainous watersheds with steep slopes.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.52 no.2
• /
• pp.96-109
• /
• 2024

The purpose of this study is to propose a method for evaluating the similarity of Show gardens using Deep Learning models, specifically VGG-16 and ResNet50. A model for judging the similarity of show gardens based on VGG-16 and ResNet50 models was developed, and was referred to as DRG (Deep Recognition of similarity in show Garden design). An algorithm utilizing GAP and Pearson correlation coefficient was employed to construct the model, and the accuracy of similarity was analyzed by comparing the total number of similar images derived at 1st (Top1), 3rd (Top3), and 5th (Top5) ranks with the original images. The image data used for the DRG model consisted of a total of 278 works from the Le Festival International des Jardins de Chaumont-sur-Loire, 27 works from the Seoul International Garden Show, and 17 works from the Korea Garden Show. Image analysis was conducted using the DRG model for both the same group and different groups, resulting in the establishment of guidelines for assessing show garden similarity. First, overall image similarity analysis was best suited for applying data augmentation techniques based on the ResNet50 model. Second, for image analysis focusing on internal structure and outer form, it was effective to apply a certain size filter (16cm × 16cm) to generate images emphasizing form and then compare similarity using the VGG-16 model. It was suggested that an image size of 448 × 448 pixels and the original image in full color are the optimal settings. Based on these research findings, a quantitative method for assessing show gardens is proposed and it is expected to contribute to the continuous development of garden culture through interdisciplinary research moving forward.

• The Korean Journal of Applied Statistics
• /
• v.23 no.1
• /
• pp.151-165
• /
• 2010

Images are often corrupted by impulse noise due to a noise sensor or channel transmission errors. The filter based on SVM(Support Vector Machine) and the improved adaptive median filtering is proposed to preserve image details while suppressing impulse noise for image restoration. Our approach uses an SVM impulse detector to judge whether the input pixel is noise. If a pixel is detected as a noisy pixel, the improved adaptive median filter is used to replace it. To demonstrate the performance of the proposed filter, extensive simulation experiments have been conducted under both salt-and-pepper and random-valued impulse noise models to compare our method with many other well known filters in the qualitative measure and quantitative measures such as PSNR and MAE. Experimental results indicate that the proposed filter performs significantly better than many other existing filters.

• Korean Institute of Interior Design Journal
• /
• v.22 no.6
• /
• pp.40-48
• /
• 2013

Utilization of CCTV in crime prevention for public safety is accepted as the most effective measure in terms of crime prevention and control. Also, it is frequently used as a device that shows evidence of an unexpected situation or record on public social relationship. However, it is rare to find a study that qualitatively accessed the monitoring performance of a certain space depending on the choice and positioning of CCTVs. Thus, this study suggested a technology that can quantitatively compare and assess the monitoring performance of CCTVs depending on view angle and effective sight range of cameras as well as the monitoring performance depending on positioning measures. For the analysis, the concept of 3-dimensional surveillance field in the form of a frustum was suggested while deriving 3-dimensional range of sight and quantitative monitoring performance by applying Isovist theory. For the analysis technology, space of analysis subject, point of view (camera), and target point (measurement node) were installed at a 3-dimensional space and in use of ray-tracing algorithm, the line segment that was visually connected between the point of view and target point was extracted and accumulated. For such verification, analysis application was constructed and then applied to four alternative models on view angle and distance as well as four alternatives on positioning in order to verify its efficacy. Through the experiment, it was possible to compare and assess visibility depending on alternatives while quantifying the results by understanding the shadow areas beyond the monitoring range.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.16 no.1
• /
• pp.176-182
• /
• 2012

Recently, a new theoretical and methodological approach for computational science is becoming more and more popular for analyzing and solving scientific problems in various scientific disciplines and applied research. Computational science is a field of study concerned with constructing mathematical models and quantitative analysis techniques and using large computing resources to solve the problems which are difficult to approach in a physical experimentally. In this paper, we present R&D of EDISON open integration platform that allows anyone like professors, researchers, industrial workers, students etc to upload their advanced research result such as simulation SW to use and share based on the cyber infrastructure of supercomputer and network. EDISON platform, which consists of 3 tiers (EDISON application framework, EDISON middleware, and EDISON infra resources) provides Web portal for education and research in 5 areas (CFD, Chemistry, Physics, Structural Dynamics, Computational Design) and user service.

• Journal of Internet Computing and Services
• /
• v.13 no.2
• /
• pp.87-98
• /
• 2012

Understanding agent's intent is an essential component of the human-computer interaction of ubiquitous computing. Because correct inference of subject's intention in ubiquitous computing system helps particularly to understand situations that involve collaboration among multiple agents or detection of situations that can pose a particular activity. This paper, inspired by people have a mechanism for interpreting one another's actions and for inferring the intentions and goals that underlie action, proposes an approach that allows a computing system to quickly recognize the intent of agents based on experience data acquired through prior capabilities of activities recognition. To proceed intention recognition, proposed method uses formulations of Hidden Markov Models (HMM) to model a system's prior experience and agents' action change, then makes for system infer intents in advance before the agent's actions are finalized while taking the perspective of the agent whose intent should be recognized. Quantitative validation of experimental results, while presenting an accurate rate, an early detection rate and a correct duration rate with detecting the intent of several people performing various activities, shows that proposed research contributes to implement effective intent recognition system.

• The Korean Journal of Physiology and Pharmacology
• /
• v.17 no.6
• /
• pp.517-523
• /
• 2013

Naphthyridine compounds are important, because they exhibit various biological activities including anticancer, antimicrobial, and anti-inflammatory activity. Some naphthyridines have antimitotic effects or demonstrate anticancer activity by inhibiting topoisomerase II. These compounds have been investigated as potential anticancer agents, and several compounds are now part of clinical trials. A series of naphthyridine derivatives were evaluated for their in vitro cytotoxic activities against human cervical cancer (HeLa), leukemia (HL-60), and prostate cancer (PC-3) cell lines using an MTT assay. Some compounds (14, 15, and 16) were more potent than colchicine against all three human cancer cell lines and compound (16) demonstrated potency with $IC_{50}$ values of 0.7, 0.1, and $5.1{\mu}M$, respectively. Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were used for quantitative structure-activity relationship (QSAR) molecular modeling of these compounds. We obtained accurate and predictive three-dimensional QSAR (3D-QSAR) models as indicated by the high PLS parameters of the HeLa ($q^2$, 0.857; $r^2$, 0.984; $r^2\;_{pred}$, 0.966), HL-60 ($q^2$, 0.777; $q^2$, 0.937; $r^2\;_{pred}$, 0.913), and PC-3 ($q^2$, 0.702; $q^2$, 0.983; $r^2\;_{pred}$, 0.974) cell lines. The 3D-QSAR contour maps suggested that the C-1 NH and C-4 carbonyl group of the naphthyridine ring and the C-2 naphthyl ring were important for cytotoxicity in all three human cancer cell lines.

• Asian-Australasian Journal of Animal Sciences
• /
• v.20 no.11
• /
• pp.1625-1630
• /
• 2007

A simulation study was conducted to evaluate the effect of reciprocal cross on the detection and characterization of Mendelian QTL in $F_2$ QTL swine populations. Data were simulated under two different mating designs. In the one-way cross design, six $F_0$ grand sires of one breed and 30 $F_0$ grand dams of another breed generated 10 $F_1$ offspring per dam. Sixteen $F_1$ sires and 64 $F_1$ dams were randomly chosen to produce a total of 640 $F_2$ offspring. In the reciprocal design, three $F_0$ grand sires of A breed and 15 $F_0$ grand dams of B breed were mated to generate 10 $F_1$ offspring per dam. Eight $F_1$ sires and 32 $F_1$ dams were randomly chosen to produce 10 $F_2$ offspring per $F_1$ dam, for a total of 320 $F_2$ offspring. Another mating set comprised three $F_0$ grand sires of B breed and 15 $F_0$ grand dams of A breed to produce the same number of $F_1$ and $F_2$ offspring. A chromosome of 100 cM was simulated with large, medium or small QTL with fixed, similar, or different allele frequencies in parental breeds. Tests between Mendelian models allowed QTL to be characterized as fixed (LC QTL), or segregating at similar (HS QTL) or different (CB QTL) frequencies in parental breeds. When alternate breed alleles segregated in parental breeds, a greater proportion of QTL were classified as CB QTL and estimates of QTL effects for the CB QTL were more unbiased and precise in the reciprocal cross than in the one-way cross. This result suggests that reciprocal cross design allows better characterization of Mendelian QTL in terms of allele frequencies in parental breeds.

• Asian-Australasian Journal of Animal Sciences
• /
• v.23 no.3
• /
• pp.333-339
• /
• 2010

The objectives of this study were to use the NRC-2001 model and DVE/OEB system to model potential nutrient supply to ruminants and to compare frost damaged (also called "frozen" wheat with normal wheat. Quantitative predictions were made in terms of: i) Truly absorbed rumen synthesized microbial protein in the small intestine; ii) Truly absorbed rumen undegraded feed protein in the small intestine; iii) Endogenous protein in the digestive tract; iv). Total truly absorbed protein in the small intestine; and v). Protein degraded balance. The overall yield losses of the frozen wheat were 24%. Results showed that using the DVE/OEB system to predict the potential nutrient supply, the frozen wheat had similar truly absorbed rumen synthesized microbial protein (65 vs. 66 g/kg DM; p>0.05), tended to have lower truly absorbed rumen undegraded feed protein (39 vs. 53 g/kg DM; p<0.10) and had higher endogenous protein (14 vs. 9 g/kg DM; p<0.05). Total truly absorbed protein in the small intestine was significantly lower (89 vs. 110 g/kg DM, p<0.05) in the frozen wheat. The protein degraded balance was similar and both were negative (-2 vs. -1 g/kg DM). Using the NRC-2001 model to predict the potential nutrient supply, the frozen wheat also had similar truly absorbed rumen synthesized microbial protein (average 56 g/kg DM; p>0.05), tended to have lower truly absorbed rumen undegraded feed protein (35 vs. 48, g/kg DM; p<0.10) and had similar endogenous protein (average 4 g/kg DM; p>0.05). Total truly absorbed protein in the small intestine was significantly lower (95 vs. 108 g/kg DM, p<0.05) in the frozen wheat. The protein degraded balance was not significantly different and both were negative (-16 vs. -19 g/kg DM). In conclusion, both models predict lower protein value and negative protein degraded balance in the frozen wheat. The frost damage to the wheat reduced nutrient content and availability and thus reduced nutrient supply to ruminants by around 12 to 19%.