• Journal of Intelligence and Information Systems
• /
• v.24 no.4
• /
• pp.219-240
• /
• 2018

Sentiment analysis, which is one of the text mining techniques, is a method for extracting subjective content embedded in text documents. Recently, the sentiment analysis methods have been widely used in many fields. As good examples, data-driven surveys are based on analyzing the subjectivity of text data posted by users and market researches are conducted by analyzing users' review posts to quantify users' reputation on a target product. The basic method of sentiment analysis is to use sentiment dictionary (or lexicon), a list of sentiment vocabularies with positive, neutral, or negative semantics. In general, the meaning of many sentiment words is likely to be different across domains. For example, a sentiment word, 'sad' indicates negative meaning in many fields but a movie. In order to perform accurate sentiment analysis, we need to build the sentiment dictionary for a given domain. However, such a method of building the sentiment lexicon is time-consuming and various sentiment vocabularies are not included without the use of general-purpose sentiment lexicon. In order to address this problem, several studies have been carried out to construct the sentiment lexicon suitable for a specific domain based on 'OPEN HANGUL' and 'SentiWordNet', which are general-purpose sentiment lexicons. However, OPEN HANGUL is no longer being serviced and SentiWordNet does not work well because of language difference in the process of converting Korean word into English word. There are restrictions on the use of such general-purpose sentiment lexicons as seed data for building the sentiment lexicon for a specific domain. In this article, we construct 'KNU Korean Sentiment Lexicon (KNU-KSL)', a new general-purpose Korean sentiment dictionary that is more advanced than existing general-purpose lexicons. The proposed dictionary, which is a list of domain-independent sentiment words such as 'thank you', 'worthy', and 'impressed', is built to quickly construct the sentiment dictionary for a target domain. Especially, it constructs sentiment vocabularies by analyzing the glosses contained in Standard Korean Language Dictionary (SKLD) by the following procedures: First, we propose a sentiment classification model based on Bidirectional Long Short-Term Memory (Bi-LSTM). Second, the proposed deep learning model automatically classifies each of glosses to either positive or negative meaning. Third, positive words and phrases are extracted from the glosses classified as positive meaning, while negative words and phrases are extracted from the glosses classified as negative meaning. Our experimental results show that the average accuracy of the proposed sentiment classification model is up to 89.45%. In addition, the sentiment dictionary is more extended using various external sources including SentiWordNet, SenticNet, Emotional Verbs, and Sentiment Lexicon 0603. Furthermore, we add sentiment information about frequently used coined words and emoticons that are used mainly on the Web. The KNU-KSL contains a total of 14,843 sentiment vocabularies, each of which is one of 1-grams, 2-grams, phrases, and sentence patterns. Unlike existing sentiment dictionaries, it is composed of words that are not affected by particular domains. The recent trend on sentiment analysis is to use deep learning technique without sentiment dictionaries. The importance of developing sentiment dictionaries is declined gradually. However, one of recent studies shows that the words in the sentiment dictionary can be used as features of deep learning models, resulting in the sentiment analysis performed with higher accuracy (Teng, Z., 2016). This result indicates that the sentiment dictionary is used not only for sentiment analysis but also as features of deep learning models for improving accuracy. The proposed dictionary can be used as a basic data for constructing the sentiment lexicon of a particular domain and as features of deep learning models. It is also useful to automatically and quickly build large training sets for deep learning models.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2004.08a
• /
• pp.157-179
• /
• 2004

The application of magnetic method for archaeoprospection has been carried out through two archaeological areas in Egypt, Abydos and Abu Sir, In order to find out tile ancient Egyptian archaeological features. The magnetic work at the selected archaeological site of Abydos area was carried out by gradiometer survey, while magnetic work at the selected archaeological site of Abu Sir area was carried out by gradiometer survey and magnetic susceptibility measurements. A gradiometer survey with raster of 0.5 m/0.5 m has been carried out on a surface area of $9600 m^2$ at Abydos area to relocate the buried Solar Boats. The magnetic data were processed using Geoplot software to treat the field noises and enhance the quality of the obtained images. The final magnetic images indicate the existence of 12 Solar Boats as well as tombs, remains of ancient rooms and walls. All of them are expected to belong to the Middle Kingdom, particularly from the 18th to 20th Dynasties. Two magnetic tools have been applied over a selected site of $25600 m^2$ at Abu Sir area in order to detect the hidden archaeological features nearby the Sun Temple. The acquisition of the magnetic data was initiated by the measurements of the topsoil magnetic susceptibility of 272 samples collected from the whole studied area, and then followed by the gradiometer survey to measure tile vertical gradient of the geomagnetic field over an area of $14400 m^2$. The magnetic susceptibility results show the presence of high concentration at the middle part of the study area with a little extension to the south western side, with maximum value of about $36{\times}10^5$ SI. They may indicate the proximity of ritual monuments. Also, they offered the site of interest for carrying out a gradiometer survey. The gradiometer results show tile existence of numerous distributed archaeological features made of mud-bricks with different shapes and sizes. They may indicate tombs, burial rooms, dissected walls; all of them are expected to belong to the 5th Dynasty of pharaohs, who used to build their buildings by mud bricks. The depth of the expected buried archaeological features has been estimated from tihe gradiometer. It is around 1.2m for deep features and 0.42 m for shallow features.

• Geomechanics and Engineering
• /
• v.16 no.4
• /
• pp.399-413
• /
• 2018

This paper addresses the issue of field measurement of excavation damage zone (EDZ) and its numerical simulation method considering both excavation unloading and blasting load effects. Firstly, a 2000 m-deep rock cavern in China is focused. A detailed analysis is conducted on the field measurement data regarding the mechanical response of rock masses subjected to excavation and blasting operation. The extent of EDZ is revealed 3.6 m-4.0 m, accounting for 28.6% of the cavern span, so it is significantly larger than rock caverns at conventional overburden depth. The rock mass mechanical response subjected to excavation and blasting is time-independent. Afterwards, based on findings of the field measurement data, a numerical evaluation method for EDZ determination considering both excavation unloading and blasting load effects is presented. The basic idea and general procedures are illustrated. It features a calibration operation of damage constant, which is defined in an elasto-plastic damage constitutive model, and a regression process of blasting load using field blasting vibration monitoring data. The numerical simulation results are basically consistent with the field measurement results. Further, some issues regarding the blasting loads, applicability of proposed numerical method, and some other factors are discussed. In conclusion, the field measurement data collected from the 2000 m-deep rock cavern and the corresponding findings will broaden the understanding of tunnel behavior subjected to excavation and blasting at great depth. Meanwhile, the presented numerical simulation method for EDZ determination considering both excavation unloading and blasting load effects can be used to evaluate rock caverns with similar characteristics.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.28 no.5
• /
• pp.1197-1207
• /
• 2018

PowerShell is command line shell and scripting language, built on the .NET framework, and it has several advantages as an attack tool, including built-in support for Windows, easy code concealment and persistence, and various pen-test frameworks. Accordingly, malwares using PowerShell are increasing rapidly, however, there is a limit to cope with the conventional malware detection technique. In this paper, we propose an improved monitoring method to observe commands executed in the PowerShell and a deep learning based malware classification model that extract features from commands using Convolutional Neural Network(CNN) and send them to Recurrent Neural Network(RNN) according to the order of execution. As a result of testing the proposed model with 5-fold cross validation using 1,916 PowerShell-based malwares collected at malware sharing site and 38,148 benign scripts disclosed by an obfuscation detection study, it shows that the model effectively detects malwares with about 97% True Positive Rate(TPR) and 1% False Positive Rate(FPR).

• Korean Journal of Cognitive Science
• /
• v.16 no.1
• /
• pp.53-67
• /
• 2005

Present study identified categorization pattern differences between experts and novices and examined whether categorization training has positive effects on problem solving. In experiment I, we examined categorization differences between groups according to expertise using mathematical equation problems. Experts classified problems based on deep structure related to problem solution methods whereas novices classified problems based on surface features. However, in the labeled categorization condition, novices' categorization pattern was not different from experts'. These results suggest that novices have difficulty identifying deep structure of problems. In experiment 2, we examined whether categorization training showing subjects deep structure of problems explicitly increases transfer performance. The results showed that solution training was more effective to expert group whereas categorization training was more effective to novice group. We have discussed that different training methods should be applied according to expertise.

• Journal of Digital Contents Society
• /
• v.18 no.1
• /
• pp.183-190
• /
• 2017

This paper proposes an audio event detection method using Deep Neural Networks (DNN). The proposed method applies Feed Forward Neural Network (FFNN) to generate output probabilities of twenty audio events for each frame. Mel scale filter bank (FBANK) features are extracted from each frame, and its five consecutive frames are combined as one vector which is the input feature of the FFNN. The output layer of FFNN produces audio event probabilities for each input feature vector. More than five consecutive frames of which event probability exceeds threshold are detected as an audio event. An audio event continues until the event is detected within one second. The proposed method achieves as 71.8% accuracy for 20 classes of the UrbanSound8K and the BBC Sound FX dataset.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1250-1256
• /
• 2019

Although one can easily generate real-world 3D mesh data using a 3D printer or a depth camera, the generated data inevitably includes unnecessary noise. Therefore, mesh denoising is essential to obtain intact 3D mesh data. However, conventional mathematical denoising methods require preprocessing and often eliminate some important features of the 3D mesh. To address this problem, this paper proposes a deep learning based 3D mesh denoising method. Specifically, we propose a convolution-based autoencoder model consisting of an encoder and a decoder. The convolution operation applied to the mesh data performs denoising considering the relationship between each vertex constituting the mesh data and the surrounding vertices. When the convolution is completed, a sampling operation is performed to improve the learning speed. Experimental results show that the proposed autoencoder model produces faster and higher quality denoised data than the conventional methods.

• The Bulletin of The Korean Astronomical Society
• /
• v.37 no.1
• /
• pp.36.2-36.2
• /
• 2012

The aim of this study was to explore the role of galaxy mergers on the formation and evolution of galaxies in galaxy clusters. For this purpose, u', g', r' deep optical imaging and multi-object spectroscopic observation were done for four rich Abell clusters at z ${\lesssim}$ 0.1 (A119, A2670, A3330, and A389) with a MOSAIC 2 CCD and Hydra spectrograph mounted on a Blanco 4-m telescope at CTIO. With the deep images, we found that about 25% of the bright red-sequence galaxies exhibited post-merger signatures in a cluster environment. This fraction was much higher than what was expected from the results of the field environment (-35%, van Dokkum 2005) and significantly low on-going merger fractions (about one-fifth of the field) appeared in the clusters currently. Taking advantage of the most up-to-date semi-analytic model, the results indicate that most of the post-merger galaxies may have carried over their merger features from their previous halo environment. All the brightest cluster galaxies in our cluster samples revealed faint structures in their halos as well as multiple nuclei in their centers seen in the deep optical images. We suggest that the mass of the BCGs increased mainly though major mergers at recent epochs based on their post-merger signatures and the large gaps in the total magnitudes between the BCGs and the second-rank BCGs. A UV bright tidal tail and tidal dwarf galaxy (TDG) candidates around the post-merger galaxy, NGC 4922, were discovered in the outskirts of the Coma cluster using the GALEX UV data. We did two-component stellar population modeling for the TDG candidates and the results indicate that they are an early form of dwarf galaxies frequently found around massive early-type galaxies in clusters. In conclusion, we suggest that the mergers of galaxies are an important driving force behind galaxy formation and evolution in cluster environments even until recent epochs.

• Journal of radiological science and technology
• /
• v.43 no.5
• /
• pp.397-404
• /
• 2020

Coronavirus disease(COVID-19) is highly infectious disease that directly affects the lungs. To observe the clinical findings from these lungs, the Chest Radiography(CXR) can be used in a fast manner. However, the diagnostic performance via CXR needs to be improved, since the identifying these findings are highly time-consuming and prone to human error. Therefore, Artificial Intelligence(AI) based tool may be useful to aid the diagnosis of COVID-19 via CXR. In this study, we explored various Deep learning(DL) approach to classify COVID-19, other viral pneumonia and normal. For the original dataset and lung-segmented dataset, the pre-trained AlexNet, SqueezeNet, ResNet18, DenseNet201 were transfer-trained and validated for 3 class - COVID-19, viral pneumonia, normal. In the results, AlexNet showed the highest mean accuracy of 99.15±2.69% and fastest training time of 1.61±0.56 min among 4 pre-trained neural networks. In this study, we demonstrated the performance of 4 pre-trained neural networks in COVID-19 diagnosis with CXR images. Further, we plotted the class activation map(CAM) of each network and demonstrated that the lung-segmentation pre-processing improve the performance of COVID-19 classifier with CXR images by excluding background features.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.31 no.2
• /
• pp.131-135
• /
• 2004

Rett syndrome is a neurological disorder seen almost exclusively in females, and found in a variety of racial and ethnic groups worldwide. First described by Dr. Andreas Rett in 1983. The specific features of the Rett syndrome is apraxia. Most of the Rett syndrome has been diagnosed erroneously to autism, cerebral palsy, and unknown developmental disorders. The etiology of the Rett syndrome is not figured out exactly but it seem to have relation with genetic factors. In this case the patient with Rett syndrome had a chief complaint of the injury of palate due to deep bite. We report this case for the satisfactory result using the bite plane to decrease the injury of the palate due to deep bite.