• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Intelligence and Information Systems
• /
• v.23 no.3
• /
• pp.69-94
• /
• 2017

Recently, increase of demand for big data analysis has been driving the vigorous development of related technologies and tools. In addition, development of IT and increased penetration rate of smart devices are producing a large amount of data. According to this phenomenon, data analysis technology is rapidly becoming popular. Also, attempts to acquire insights through data analysis have been continuously increasing. It means that the big data analysis will be more important in various industries for the foreseeable future. Big data analysis is generally performed by a small number of experts and delivered to each demander of analysis. However, increase of interest about big data analysis arouses activation of computer programming education and development of many programs for data analysis. Accordingly, the entry barriers of big data analysis are gradually lowering and data analysis technology being spread out. As the result, big data analysis is expected to be performed by demanders of analysis themselves. Along with this, interest about various unstructured data is continually increasing. Especially, a lot of attention is focused on using text data. Emergence of new platforms and techniques using the web bring about mass production of text data and active attempt to analyze text data. Furthermore, result of text analysis has been utilized in various fields. Text mining is a concept that embraces various theories and techniques for text analysis. Many text mining techniques are utilized in this field for various research purposes, topic modeling is one of the most widely used and studied. Topic modeling is a technique that extracts the major issues from a lot of documents, identifies the documents that correspond to each issue and provides identified documents as a cluster. It is evaluated as a very useful technique in that reflect the semantic elements of the document. Traditional topic modeling is based on the distribution of key terms across the entire document. Thus, it is essential to analyze the entire document at once to identify topic of each document. This condition causes a long time in analysis process when topic modeling is applied to a lot of documents. In addition, it has a scalability problem that is an exponential increase in the processing time with the increase of analysis objects. This problem is particularly noticeable when the documents are distributed across multiple systems or regions. To overcome these problems, divide and conquer approach can be applied to topic modeling. It means dividing a large number of documents into sub-units and deriving topics through repetition of topic modeling to each unit. This method can be used for topic modeling on a large number of documents with limited system resources, and can improve processing speed of topic modeling. It also can significantly reduce analysis time and cost through ability to analyze documents in each location or place without combining analysis object documents. However, despite many advantages, this method has two major problems. First, the relationship between local topics derived from each unit and global topics derived from entire document is unclear. It means that in each document, local topics can be identified, but global topics cannot be identified. Second, a method for measuring the accuracy of the proposed methodology should be established. That is to say, assuming that global topic is ideal answer, the difference in a local topic on a global topic needs to be measured. By those difficulties, the study in this method is not performed sufficiently, compare with other studies dealing with topic modeling. In this paper, we propose a topic modeling approach to solve the above two problems. First of all, we divide the entire document cluster(Global set) into sub-clusters(Local set), and generate the reduced entire document cluster(RGS, Reduced global set) that consist of delegated documents extracted from each local set. We try to solve the first problem by mapping RGS topics and local topics. Along with this, we verify the accuracy of the proposed methodology by detecting documents, whether to be discerned as the same topic at result of global and local set. Using 24,000 news articles, we conduct experiments to evaluate practical applicability of the proposed methodology. In addition, through additional experiment, we confirmed that the proposed methodology can provide similar results to the entire topic modeling. We also proposed a reasonable method for comparing the result of both methods.

• Journal of the Speleological Society of Korea
• /
• no.68
• /
• pp.23-73
• /
• 2005

Purpose of study; The purpose of this study is specifically classified as two parts. The one is to attempt the chronological annals of Quaternary topographic surface through the study over the formation process of alluvial surfaces in our country, setting forth the alluvial surfaces lower-parts of Han River area, as the basic deposit, and comparing it to the marginal landform surfaces. The other is to attempt the classification of micro morphology based on the and condition premising the land use as a link for the regional development in the lower-parts of Han river area. Reasons why selected the Lower-parts of Han river area as study objects: 1. The change of river course in this area is very serve both in vertical and horizontal sides. With a situation it is very easy to know about the old geography related to the formation process of topography. 2. The component materials of gravel, sand, silt and clay are deposited in this area. Making it the available data, it is possible to consider about not oかy the formation process of topography but alsoon the development history to some extent. 3. The earthen vessel, a fossil shell fish, bone, cnarcoal and sea-weed are included in the alluvial deposition in this area. These can be also valuable data related to the chronological annals. 4. The bottom set conglometate beds is also included in the alluvial deposits. This can be also valuable data related to the research of geomorphological development. 5. Around of this area the medium landform surface, lower landform surface, pediment and basin, are existed, and these enable the comparison between the erosion surfaces and the alluvial surfaces. Approach : 1. Referring to the change of river beds, I have calculated the vertical and horizontal differences comparing the topographic map published in 1916 with that published in 1966 and through the field work 2. In classifying the landform, I have applied the method of micro morphological classification in accordance with the synthetic index based upon the land conditions, and furthermore used the classification method comparing the topographic map published in 1916 and in that of 1966. 3. I have accorded this classification with the classification by mapping through appliying the method of classification in the development history for the field work making the component materials as the available data. 4. I have used the component materials, which were picked up form the outcrop of 10 places and bored at 5 places, as the available data. 5. I have referred to Hydrological survey data of the ministry of Construction (since 1916) on the overflow of Han-river, and used geologic map of Seoul metropolitan area. Survey Data, and general map published in 1916 by the Japanese Army Survbey Dept., and map published in 1966 by the Construction Research Laboratory and ROK Army Survey Dept., respectively. Conclusion: 1. Classification of Morphology: I have added the historical consideration for development, making the component materials and fossil as the data, to the typical consideration in accordance with the map of summit level, reliefe and slope distribution. In connection with the erosion surface, I have divided into three classification such as high, medium and low-,level landform surfaces which were classified as high and low level landform surfaces in past. furthermore I have divided the low level landform surface two parts, namely upper-parts(200-300m) and bellow-parts(${\pm}100m$). Accordingly, we can recognize the three-parts of erosion surface including the medium level landform surface (500-600m) in this area. (see table 22). In condition with the alluvial surfaces I have classified as two landform surfaces (old and new) which was regarded as one face in past. Meamwhile, under the premise of land use, the synthetic, micro morphological classification based upon the land condition is as per the draw No. 19-1. This is the quite new method of classification which was at first attempted in this country. 2. I have learned that the change of river was most severe at seeing the river meandering rate from Dangjung-ni to Nanjido. As you seee the table and the vertical and horizontal change of river beds is justly proportionable to the river meandering rate. 3. It can be learned at seeing the analysis of component materials of alluvial deposits that the component from each other by areas, however, in the deposits relationship upper stream, and between upper parts and below parts I couldn't always find out the regular ones. 4. Having earthern vessel, shell bone, fossil charcoal and and seaweeds includen in the component materials such as gravel, clay, sand and silt in Dukso and Songpa deposits area. I have become to attempt the compilation of chronicle as yon see in the table 22. 5. In according to hearing of basemen excavation, the bottom set conglomerate beds of Dukso beds of Dukso-beds is 7m and Songpa-beds is 10m. In according to information of dredger it is approx. 20m in the down stream. 6. Making these two beds as the standard beds, I have compared it to other beds. 7 The coarse sand beds which is covering the clay-beds of Dukso-beds and Nanjidobeds is shown the existence of so-called erosion period which formed the gap among the alluvial deposits of stratum. The former has been proved by the sorting, bedding and roundness which was supplied by the main stream and later by the branch stream, respectively. 8. If the clay-beds of Dukeo-bed and Songpa-bed is called as being transgressive overlap, by the Eustatic movement after glacial age, the bottom set conglomerate beds shall be called as being regressive overlap at the holocene. This has the closest relationship with the basin formation movement of Seoul besides the Eustatic movement. 9. The silt-beds which is the main component of deposits of flood plain, is regarded as being deposited at the Holocene in the comb ceramic and plain pottery ages. This has the closest relationship with the change of river course and river beds.

• Economic and Environmental Geology
• /
• v.52 no.6
• /
• pp.541-554
• /
• 2019

The Taebaeksan Zone of the Okcheon Belt is a prominent fold-thrust belt, preserving evidence for overlapped polyphase and diachronous orogenic events during crustal evolution of the Korean Peninsula. The Pyeongchang-Jeongseon area of the northwestern Taebaeksan Zone is fault-bounded on the western Jucheon and southern Yeongwol areas, showing lateral variations in stratigraphy and structural geometries. For better understanding these geological characteristics of the northwestern Taebaeksan Zone, we have studied the structural geometry of the Pyeongchang-Jeongseon area. For this, we have firstly carried out the SHRIMP U-Pb age analysis of the age-unknown sedimentary rock to clarify stratigraphy for structural interpretation. The results show the late Carboniferous to middle Permian dates, indicating that it is correlated to the Upper Paleozoic Pyeongan Supergroup. In addition to this, we interpreted the geometric relationships between structural elements from the detailed field investigation of the study area. The major structure of the northwestern Taebaeksan Zone is the regional-scale Jeongseon Great syncline, having NE-trending hinge with second-order folds such as the Jidongri and Imhari anticlines and the Nambyeongsan syncline. Based on the stereographic and down-plunge projections of the structureal elements, the structural geometry of the Jeongseon Great syncline can be interpreted as a synformal culmination, plunging slightly to the south at its southern area, and north at the northern area. The different map patterns of the northern and southern parts of the study area should be resulted in different erosion levels caused by the plunging hinges. Considering the Jeongseon Great syncline is the major structure that constrains the distribution of the Paleozoic strata of the Pyeongchang and Jeongseon areas, the symmetric repetition of the lower Paleozoic Joseon Supergroup in both limbs should be re-examined by structural mapping of the Hangmae and Hoedongri formations in the Pyeongchang and Jeongseon areas.

• Journal of Korean Society of Transportation
• /
• v.19 no.4
• /
• pp.35-47
• /
• 2001

Traffic represents one of the largest sources of primary air pollutants in urban area. As a consequence. numerous abatement strategies are being pursued to decrease the ambient concentration of pollutants. A characteristic of most of the these strategies is a requirement for accurate data on both the quantity and spatial distribution of emissions to air in the form of an atmospheric emission inventory database. In the case of traffic pollution, such an inventory must be compiled using activity statistics and emission factors for vehicle types. The majority of inventories are compiled using passive data from either surveys or transportation models and by their very nature tend to be out-of-date by the time they are compiled. The study of current trends are towards integrating urban traffic control systems and assessments of the environmental effects of motor vehicles. In this study, a methodology of motor vehicle emission calculation by using real-time traffic data was studied. A methodology for estimating emissions of CO at a test area in Seoul. Traffic data, which are required on a street-by-street basis, is obtained from induction loops of traffic control system. It was calculated speed-related mass of CO emission from traffic tail pipe of data from traffic system, and parameters are considered, volume, composition, average velocity, link length. And, the result was compared with that of a method of emission calculation by VKT(Vehicle Kilometer Travelled) of vehicles of category.

• Korean Journal of Remote Sensing
• /
• v.29 no.5
• /
• pp.517-525
• /
• 2013

Time-series data of Normal Difference Vegetation Index (NDVI) obtained by the Moderate-resolution Imaging Spectroradiometer(MODIS) satellite imagery gives a waveform that reveals the characteristics of the phenology. The waveform can be decomposed into harmonics of various periods by the Fourier transformation. The resulting $n^{th}$ harmonics represent the amount of NDVI change in a period of a year divided by n. The values of each harmonics or their relative relation have been used to classify the vegetation species and to build a vegetation map. Here, we propose a method to estimate the annual amount of carbon absorbed on the forest from the $1^{st}$ harmonic NDVI value. The $1^{st}$ harmonic value represents the amount of growth of the leaves. By the allometric equation of trees, the growth of leaves can be considered to be proportional to the total amount of carbon absorption. We compared the $1^{st}$ harmonic NDVI values of the 6220 sample points with the reference data of the carbon absorption obtained by the field survey in the forest of South Korea. The $1^{st}$ harmonic values were roughly proportional to the amount of carbon absorption irrespective of the species and ages of the vegetation. The resulting proportionality constant between the carbon absorption and the $1^{st}$ harmonic value was 236 tCO2/5.29ha/year. The total amount of carbon dioxide absorption in the forest of South Korea over the last ten years has been estimated to be about 56 million ton, and this coincides with the previous reports obtained by other methods. Considering that the amount of the carbon absorption becomes a kind of currency like carbon credit, our method is very useful due to its generality.

• Journal of the Society of Cosmetic Scientists of Korea
• /
• v.43 no.1
• /
• pp.61-68
• /
• 2017

The key quality attributes of the pressed powder, one of base makeup products, are skin-affinity and spreadability. In general, there was a limit to meet skin-affinity and spreadability simultaneously, which are opposite attributes each other. In this study, air jet mill process was tried to satisfy two main properties. Skin-affinity was improved by a wet coating of sericite with a mixture of lauroyl lysine (LL) and sodium cocoyl glutamate (SCG). The application of mono-dispersed polymethyl methacrylate (PMMA) and diphenyl dimethicone/vinyl diphenyl dimethicone/silsesquioxane crosspolymer (DDVDDSC) improved both qualities. Air jet mill process has been mainly applied in the pharmaceutical and food industries, and is a method used for processing powder materials in cosmetic field. In this study, we were able to complete makeup cosmetics with an optimum particle size $6.8{\mu}m$ by combining the air jet mill process at the manufacturing stage. It was confirmed that the Ti element was uniformly distributed throughout the cosmetics by EDS mapping, and that the corners of the tabular grains were rounded by SEM analysis. It is considered that this can provide an effect of improving the spreadability when the cosmetic is applied to the skin by using a makeup tool. LL with excellent skin compatibility and SCG derived from coconut with little skin irritation were wet coated to further enhance the adhesion of sericite. SEM images were analyzed to evaluate effect of the dispersion and uniformity of PMMA on spreadability. With the spherical shapes of similar size, it was found that the spreading effect was further increased when the distribution was homogeneously mono-dispersed. The dispersion and spreadability of PMMA were confirmed by measuring the kinetic friction and optimal content was determined. The silicone rubber powder, DDVDDSC, was confirmed by evaluating the hardness, spreading value, and drop test. Finally, it was found that the dispersion of PMMA and silicone rubber powder affected spreadability. Such makeup cosmetics have excellent stability in drop test while having appropriate hardness, and good stability over time. Taken together, it is concluded that air jet mill process can be utilized as a method to improve skin-affinity and spreadability of the pressed powder.

• The Journal of the Petrological Society of Korea
• /
• v.22 no.3
• /
• pp.219-233
• /
• 2013

Spherulitic rhyolite occur as part of ring dyke which showing a vertical flowage of $60^{\circ}{\sim}90^{\circ}$, of the Jangsan cauldron was studied. The spherulites range in diameter from a few millimeters to 2.8 centimeters or more, and average 5~10 millimeters. It belongs to radiated simple spherulite type. They consist of a core of moderate brown dense material encased by a thin crust, a few millimeters thick at most of white grey material. The spherulites frequently have a radiating fibrous structure, which are thought to have formed as a consequence of rapid mineral growth caused by very fast cooling of the dykes in shallow depth near the surface. EPMA examination of the concentric-zoned core of spherulites show that they are mainly composed of cryptocrystalline-fibrous intergrowth of silica minerals and alkali feldspars which have $SiO_2$ 82% or more, $Al_2O_3$ 7~10%, $Na_2O+K_2O$ less than 8%. The feldspar compositions of the spherulites lie essentially within the sanidine field. XRD examination show that spherulites are mainly composed of quartz, sanidine, albite with minor mica, kaolinite and chlorite. According to X-ray mapping, the spherulites are enriched in $SiO_2$ in the core and partly enriched $Na_2O$ or $K_2O$, $Al_2O_3$ in the shell that reflect in compositional zoning with increasing spherulitic devitrification. The feathery and non-equant crystal shapes of spherulites from rhyolite dyke of Jangsan cauldron suggest that they may have formed during the rapid cooling of dyke under the static state, or faster velocity of devitrification from glassy materials than movement velocity of the magma intrusion. The spherulitic rhyolite originated from high-silica(75.4~75.7 wt.%) rhyolite magma.

• Journal of Soil and Groundwater Environment
• /
• v.7 no.3
• /
• pp.33-44
• /
• 2002

A field technique for assessing the transmissive fractures in an aquifer was applied to a fractured rock formation in Youngchun area Korea. Geological mapping and detailed acoustic borehole teleview(BHTV) logging were performed to obtain information about the fractures. The study area consists predominantly of two types of fractures. The fracture sets of low angle partings such as bedding and sheeting plains have strike N70-80$^{\circ}$W, 25$^{\circ}$-30$^{\circ}$SW and N3S$^{\circ}$W, 12$^{\circ}$NE, respectively. In areas of high fractures, on the other hand, the major fracture sets show strike N80$^{\circ}$W and dip 70$^{\circ}$-85$^{\circ}$SW, N10$^{\circ}$E.85$^{\circ}$SE in sedimentry rocks, N40-50$^{\circ}$E.85$^{\circ}$SE/85$^{\circ}$NE, N70$^{\circ}$E.80$^{\circ}$SE, and N7$^{\circ}$-75$^{\circ}$W.80$^{\circ}$SW in granites and volcanic rocks. Injection tests have been performed to identify discrete production zones and quantify the vertical distribution of hydraulic conductivity. The calculated hydraulic conductivities range from 3.363E-10 to 2.731E-6, showing that the difference between maximum and minimum value is four order of magnitude. Dominant section in hydraulic conductivity is extensively fractured. Geophysical logging was carried out to clarify characterization of the distribution of fracture zones. Transmissive fractures were evaluated through the comparison of the results obtained by each method. The temperature logs appeared to be a good indicator that can distinguish a high transmissive fractures from a common fractures in hydraulic conductivity. In numerous cases, evidence of fluid movement was amplified in the temperature gradient log. The fracture sets of N70-80$^{\circ}$W.60-85$^{\circ}$NE/SW N75-80$^{\circ}$W.25-30$^{\circ}$SW, N50-64$^{\circ}$W.60-85$^{\circ}$NE, N35-45$^{\circ}$E.65-75$^{\circ}$SE, and N65-72$^{\circ}$E.80$^{\circ}$SE/60$^{\circ}$NW were idenfied as a distinct transmissive fractures through the results of each tests.

• Journal of the Korean Geographical Society
• /
• v.38 no.4
• /
• pp.630-639
• /
• 2003

This study aims to analysis the distribution and change of cropland and forest, the Onseong, Saebyeol, and Eundeok counties on the lower reach of Duman(Tumen) river, northeast Korea, using 1992 year Landsat TM data, 2000 year Landsat ETM data, and digital terrain elevation data(DTED). Land cover and land use of the study areas are classified into cropland, forest, village, and water body, using the supervised classification method including 1:50,000 DTED analysis, image band composition, and principal component analysis(PCA). Results of quantitative analysis present that each growth rate of cropland of Onseong and Eundeok are 22.8% and 14.7% corresponding to decreasing rates of forest, 8% and 13.6% during 8 years from 1992 to 2000. In Onseong, Saebyeol, and Eundeok, each values of mean elevations and slope gradients increased to 192m, 95m, and 91m from 157m, 85m, and 78m, and to 6.6$^{\circ}$, 3.0$^{\circ}$, and 4.4$^{\circ}$ from 5.2$^{\circ}$, 2.5$^{\circ}$, and 3.0$^{\circ}$. Especially, in case of newly developed cropland, the values of mean elevation and mean gradient have 225m, 122m, and 127m, and 9.4$^{\circ}$, 5.1$^{\circ}$, and 8.0$^{\circ}$, in above three regions. These new croplands were developing along to deeper valleys and toward lower hill and mountain slope up to knickpoint zone of gradient change. Deforested lands for cropland have formed irregular pattern of patch-type, and become sources for the sheet erosion, rilling and gulleying in mountain slope and sedimentation in local river channel. Though there were no field checking, analysis using landsat images and GIS mapping can help understand actual environmental problems relating to cropland development of mountain slope in North Korea.