• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.27-40
• /
• 2023

With the increase in the spread of smart devices and the impact of COVID-19, the consumption of media contents through smart devices has significantly increased. Along with this trend, the amount of media contents viewed through OTT platforms is increasing, that makes contents recommendations on these platforms more important. Previous contents-based recommendation researches have mostly utilized metadata that describes the characteristics of the contents, with a shortage of researches that utilize the contents' own descriptive metadata. In this paper, various text data including titles and synopses that describe the contents were used to recommend similar contents. KLUE-RoBERTa-large, a Korean language model with excellent performance, was used to train the model on the text data. A dataset of over 20,000 contents metadata including titles, synopses, composite genres, directors, actors, and hash tags information was used as training data. To enter the various text features into the language model, the features were concatenated using special tokens that indicate each feature. The test set was designed to promote the relative and objective nature of the model's similarity classification ability by using the three contents comparison method and applying multiple inspections to label the test set. Genres classification and hash tag classification prediction tasks were used to fine-tune the embeddings for the contents meta text data. As a result, the hash tag classification model showed an accuracy of over 90% based on the similarity test set, which was more than 9% better than the baseline language model. Through hash tag classification training, it was found that the language model's ability to classify similar contents was improved, which demonstrated the value of using a language model for the contents-based filtering.

• Wind and Structures
• /
• v.36 no.4
• /
• pp.237-247
• /
• 2023

The emergence of high-rise buildings has necessitated frequent structural health monitoring and maintenance for safety reasons. Wind causes damage and structural changes on tall structures; thus, safe structures should be designed. The pressure developed on tall buildings has been utilized in previous research studies to assess the impacts of wind on structures. The wind tunnel test is a primary research method commonly used to quantify the aerodynamic characteristics of high-rise buildings. Wind pressure is measured by placing pressure sensor taps at different locations on tall buildings, and the collected data are used for analysis. However, sensors may malfunction and produce erroneous data; these data losses make it difficult to analyze aerodynamic properties. Therefore, it is essential to generate missing data relative to the original data obtained from neighboring pressure sensor taps at various intervals. This study proposes a deep learning-based, deep convolutional generative adversarial network (DCGAN) to restore missing data associated with faulty pressure sensors installed on high-rise buildings. The performance of the proposed DCGAN is validated by using a standard imputation model known as the generative adversarial imputation network (GAIN). The average mean-square error (AMSE) and average R-squared (ARSE) are used as performance metrics. The calculated ARSE values by DCGAN on the building model's front, backside, left, and right sides are 0.970, 0.972, 0.984 and 0.978, respectively. The AMSE produced by DCGAN on four sides of the building model is 0.008, 0.010, 0.015 and 0.014. The average standard deviation of the actual measures of the pressure sensors on four sides of the model were 0.1738, 0.1758, 0.2234 and 0.2278. The average standard deviation of the pressure values generated by the proposed DCGAN imputation model was closer to that of the measured actual with values of 0.1736,0.1746,0.2191, and 0.2239 on four sides, respectively. In comparison, the standard deviation of the values predicted by GAIN are 0.1726,0.1735,0.2161, and 0.2209, which is far from actual values. The results demonstrate that DCGAN model fits better for data imputation than the GAIN model with improved accuracy and fewer error rates. Additionally, the DCGAN is utilized to estimate the wind pressure in regions of buildings where no pressure sensor taps are available; the model yielded greater prediction accuracy than GAIN.

• The Korean Journal of Physiology and Pharmacology
• /
• v.27 no.3
• /
• pp.267-275
• /
• 2023

Cardiotoxicity, particularly drug-induced Torsades de Pointes (TdP), is a concern in drug safety assessment. The recent establishment of human induced pluripotent stem cell-derived cardiomyocytes (human iPSC-CMs) has become an attractive human-based platform for predicting cardiotoxicity. Moreover, electrophysiological assessment of multiple cardiac ion channel blocks is emerging as an important parameter to recapitulate proarrhythmic cardiotoxicity. Therefore, we aimed to establish a novel in vitro multiple cardiac ion channel screening-based method using human iPSC-CMs to predict the drug-induced arrhythmogenic risk. To explain the cellular mechanisms underlying the cardiotoxicity of three representative TdP high- (sotalol), intermediate- (chlorpromazine), and low-risk (mexiletine) drugs, and their effects on the cardiac action potential (AP) waveform and voltage-gated ion channels were explored using human iPSC-CMs. In a proof-of-principle experiment, we investigated the effects of cardioactive channel inhibitors on the electrophysiological profile of human iPSC-CMs before evaluating the cardiotoxicity of these drugs. In human iPSC-CMs, sotalol prolonged the AP duration and reduced the total amplitude (TA) via selective inhibition of I_Kr and I_Na currents, which are associated with an increased risk of ventricular tachycardia TdP. In contrast, chlorpromazine did not affect the TA; however, it slightly increased AP duration via balanced inhibition of I_Kr and I_Ca currents. Moreover, mexiletine did not affect the TA, yet slightly reduced the AP duration via dominant inhibition of I_Ca currents, which are associated with a decreased risk of ventricular tachycardia TdP. Based on these results, we suggest that human iPSC-CMs can be extended to other preclinical protocols and can supplement drug safety assessments.

• International journal of advanced smart convergence
• /
• v.12 no.3
• /
• pp.89-103
• /
• 2023

Obstructive sleep apnea (OSA) is one of the most prevalent sleep disorders that can lead to serious consequences, including hypertension and/or cardiovascular diseases, if not treated promptly. Continuous positive airway pressure (CPAP) is widely recognized as the most effective treatment for OSA, which needs the proper titration of airway pressure to achieve the most effective treatment results. However, the process of CPAP titration can be time-consuming and cumbersome. There is a growing importance in predicting personalized CPAP pressure before CPAP treatment. The primary objective of this study was to optimize the CPAP titration process for obstructive sleep apnea patients through EEG feature engineering with machine learning techniques. We aimed to identify and utilize the most critical EEG features to forecast key OSA predictive indicators, ultimately facilitating more precise and personalized CPAP treatment strategies. Here, we analyzed 126 OSA patients' PSG datasets before and after the CPAP treatment. We extracted 29 EEG features to predict the features that have high importance on the OSA prediction index which are AHI and SpO2 by applying the Shapley Additive exPlanation (SHAP) method. Through extracted EEG features, we confirmed the six EEG features that had high importance in predicting AHI and SpO2 using XGBoost, Support Vector Machine regression, and Random Forest Regression. By utilizing the predictive capabilities of EEG-derived features for AHI and SpO2, we can better understand and evaluate the condition of patients undergoing CPAP treatment. The ability to predict these key indicators accurately provides more immediate insight into the patient's sleep quality and potential disturbances. This not only ensures the efficiency of the diagnostic process but also provides more tailored and effective treatment approach. Consequently, the integration of EEG analysis into the sleep study protocol has the potential to revolutionize sleep diagnostics, offering a time-saving, and ultimately more effective evaluation for patients with sleep-related disorders.

• International Journal of Industrial Entomology and Biomaterials
• /
• v.46 no.1
• /
• pp.16-23
• /
• 2023

The silkworm's dormancy and embryonic development are accomplished through the interaction of various genes. Analysis of the expression of several interacting genes can predict the embryonic stage of silkworms. In this study, we analyzed the changes in the expression level of genes at each stage during the embryonic development of dormant silkworm eggs and selected genes that can predict the hatching time. Jam123 and Jam124 silkworms were collected after egg laying, and the silkworm eggs were preserved using a double refrigeration method and expression analysis was performed for 23 genes during embryogenesis. There were 5 genes showing significant changes during embryogenesis: UDP-glucuronosyltransferases (BmUGTs), heat shock protein hsp20.8 (BmHsp20.8), Cytochromes b5-like proteins (BmCytb5), Krüppel homolog 1 (BmKr-h1), and cuticular protein RR-1 motif 41 (BmCpr41). As a result of quantitative comparison of the expression levels of these 5 genes through real-time PCR, the BmUGTs gene showed a difference between Jam123 and Jam124, making it difficult to see it as an indicator for predicting hatching time. However, the BmHsp20.8 gene had a common expression decreased at the imminent hatching stage. In addition, it was confirmed that the expression level of the BmCytb5 gene decreased to the lowest level at the time of imminent hatching, and the expression of the BmKr-h gene was made only at the time of imminent hatching. The expression of the last BmCpr41 gene can be confirmed only at the time of imminent hatching, and it was confirmed that it shows a rapid increase right before hatching. Taken together, these results suggest that expression analysis of BmHsp20.8, BmCytb5, BmKr-h1, and BmCpr41 genes can determine the stage of embryogenesis, predict hatching time, which facilitate better management of silkworm eggs.

• Nuclear Engineering and Technology
• /
• v.25 no.1
• /
• pp.91-101
• /
• 1993

To predict accurately the thermal hydraulic behavior of light water reactors during normal or abnormal operation, the accurate estimation of the void distribution is required. Up to date, many techniques for predicting void fraction of two-phase flow systems have been suggested. Among these techniques, the drift-flux model is widely used because of its exact calculation ability and simplicity. However, to get more accurate prediction of void fraction using drift-flux model, slip and flow regime effects must be considered more properly In the drift-flux method, these two effects are accounted for by two drift-flux parameters ; $C_{o}$ and (equation omitted). At earlier stage, $C_{o}$ is measured in a circular tube. In this study, $C_{o}$ is experimentally determined by measuring local void fraction and vapor velocity distribution in a rectangular subchannel having 4 heating rods which simulates nuclear subchannels. The measurements are peformed with two-electrical conductivity probes which are known to be adequate for measuring local parameters. The experiments are performed at low flow rate and the system pressure less than 3 atmo spheric pressure. In this experiment, (equation omitted), is not measured, but quoted from well-known empirical correlation to formulate $C_{o}$. Finally, $C_{o}$ is expressed as a function of channel averaged void fraction. fraction.

• Journal of the Korean Geotechnical Society
• /
• v.24 no.6
• /
• pp.77-84
• /
• 2008

The existing equations for radial consolidation cannot account for the changes of well resistance with time and cannot predict the appropriate in-situ consolidation curve. In this study, small cylinder cell tests are performed to evaluate the discharge capacity of PVD. Also, a block sample of 1.2 m in diameter and 2.0 m in height was consolidated to observe the change in the drainage capacity with time for three types of PVD. From the test results on a block sample, the drainage curves normalized with initial drainage of each PVD are similar, regardless of the PVD type and the consolidation curve, which is predicted using solutions of radial consolidation based on the discharge capacity measured in a small cylinder cell tests, significantly overestimates the degree of consolidation. The term of well resistance in the radial consolidation solution was back-calculated to fit the consolidation curve of a large block sample and it is defined as the time dependent well resistance factor, L(t). The L(t) was found to be linearly proportional to the dimensionless time factor, Th. It was also shown that the consolidation curve evaluated by using L(t) provides more accurate prediction than the existing solution.

• Journal of the Korean Geotechnical Society
• /
• v.22 no.11
• /
• pp.55-64
• /
• 2006

It is very Important to predict the damage zone of a rock mass induced by blasting for the excavation of an underground cavity such as a tunnel, as the damage zones incur mechanical and hydraulic instability of the rock mass potentially. Complicated blasting processes that can hinder the proper characterization of the damage zone can be effectively represented by two loading mechanisms. The first mechanism is the dynamic impulsive load-generating stress waves that radiate outwards immediately after detonation. This load creates a crushed annulus along with cracks around the blasthole. The second is the gas pressure that remains for an extended time after detonation. As the gas pressure reopens some arrested cracks and extends these, it contributes to the final structure of the damage zone induced by the blasting. This paper presents a simple method to evaluate the damage zone induced by gas pressure during rock blasting. The damage zone is characterized by analyzing crack propagations from the blasthole. To do this, a model of a blasthole with a number of radial cracks that are equal in length in a homogeneous infinite elastic plane is considered. In this model, crack propagation is simulated through the use of only two conditions: a crack propagation criterion and the mass conservation of the gas. The results show that the stress intensity factor of a crack decreases as the crack propagates from the blasthole, which determines the crack length. In addition, it was found that the blasthole pressure continues to decrease during crack propagation.

• Journal of the Korean Geotechnical Society
• /
• v.25 no.6
• /
• pp.5-16
• /
• 2009

The is study is to evaluate the small strain shear modulus ($G_{max}$) of Busan clay using in-situ penetration tests. A series of dilatometer tests (DMT) and piezocone penetration tests (CPTu) are performed at Busan newport and Noksan sites, and hybrid oedometer tests are also carried out on the specimens obtained from both sites. The $G_{max}$ is evaluated from the shear wave velocity ($V_s$) measured by the bender elements installed at the boundary of oedometer cell. By analyzing these data, the relationship of $G_{max}$ and state variables, such as confined stress and void ratio, is developed. The analysis of lab and in-situ test results reveals that the ratio of $G_{max}$ to $q_t$ is inversely proportional to the plasticity index while the ratio of $G_{max}$ to $E_D$ has a linear relationship with ($I/I_D$)$(p_a/{\sigma}'_v)^{0.5}$. Two correlations suggested in this study, based on CPT and DMT results, appear to provide reasonable predictions of the small strain shear modulus.

• Journal of the Korean Geotechnical Society
• /
• v.23 no.7
• /
• pp.57-64
• /
• 2007

This study proposes a method to estimate drilling torque during ground boring with an aid of electrical energy required for rotating a boring-auger. Ground boring is commonly used in geotechnical engineering such as preboring precast pile installation, soil-cement grouting, ground exploration and so forth. In order to understand the correlation between required electrical energy to rotate the boring auger and the drilling torque, a small laboratory apparatus was designed and a pilot study was performed. The apparatus rotates common drill bits of $D=5{\sim}25mm$ in CBR specimens. The velocity of a bit is 19 RPM and predefined using a reduction gear which connects a main rotation axis to a 25 Watts AC electrical motor shaft. In the middle of drilling the motor current increments and the drilling torque were measured and the correlation between the current and the torque was obtained through linear square fits. Based on the correlation the acquired motor current during drilling was applied to predict the drilling torque in consequent testing and the prediction results were compared to the measured torque. The comparison leads a conclusion that the motor current during drilling using electrical power may be a good indicator to estimate/determine strength characteristics of the ground.