• Journal of the Korea Computer Graphics Society
• /
• v.30 no.4
• /
• pp.11-18
• /
• 2024

In partial nephrectomy for kidney cancer treatment, accurate segmentation of the kidney tumor is crucial for surgical planning, as it provides essential information on the precise size and location of the tumor. However, it is challenging due to the tumor's similar intensity to surrounding organs and the variability in its location and size across patients. In this study, we propose a hybrid network that integrates a convolutional neural network and a transformer to capture both local and global features, aiming to improve the segmentation performance of kidney tumors. We validated our method through comparative experiments with UNETR++, outperforming it with a Dice Similarity Coefficient (DSC) of 78.54% and a precision of 85.0 7%. Moreover, in the analysis by tumor size, our method demonstrated improvements by reducing over-segmentation and outlier cases observed in UNETR++.

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.1
• /
• pp.37-44
• /
• 2015

In traditional social network analysis, the betweenness centrality measure has been heavily used to identify the relative importance of nodes. Since the time complexity to calculate the betweenness centrality is very high, however, it is difficult to get it of each node in large-scale social network where there are so many nodes and edges. In our past study, we defined a new type of network, called the expanded ego network, which is built only with each node's local information, i.e., neighbor information of the node's neighbor nodes, and also defined a new measure, called the expanded ego betweenness centrality. In this paper, We propose algorithm that quickly computes expanded ego betweenness centrality by exploiting structural properties of expanded ego network. Through the experiment with virtual network used Barab$\acute{a}$si-Albert network model to represent the generic social network and facebook network to represent actual social network, We show that the node's importance rank based on the expanded ego betweenness centrality has high similarity with that the node's importance rank based on the existing betweenness centrality. We also show that the proposed algorithm computes the expanded ego betweenness centrality quickly than existing algorithm.

• Economic and Environmental Geology
• /
• v.35 no.2
• /
• pp.137-147
• /
• 2002

Natural geological and environmental processes reflect to element abundances in geological materials on the surface. This study aims to elucidate a possibility of geostatistical application to differentiate geochemical anomalies affected by anthropogenic and geogenic factors. A regional geochemical map was produced using 'inverse distance weight interpolation' method for analytical results of stream sediments «150 11m) which were collected from 2,290 first- to second-order streams over the whole Gyeonggi Province. The Jurassic granitic batholith in the southeastern province was selected as a target for the geostatistical examination. Factor analysis was conducted using 22 elements for stream sediments from 445 drainage basins over the granitic body. Co, Cr, Sc, MgO, Fe$_{2}$O$_{3}$, V, and Ni were grouped with high correlation coefficients and the depletion of the components may reflect the whole-rock chemistry of the granite. Regression analysis was done using Co, Cr, and Sc as dependent variables and other six components as independent variables, and the results were drawn as maps. The maps acquired generally show quite similar distribution patterns with those of concentrations of each variable. The similarity in the spatial patterns between the two maps indicates that the application of regression statistics can be valid for the interpretation of regional geochemical data. However, some components show local discrepancies which may be influenced by secondary factors regardless of the basement lithology. The regression analysis may be effective in extracting local geochemical anomalies which may reflect rather anthropogenic pollutions than geogenic influences.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.8 no.2
• /
• pp.45-51
• /
• 1990

DEM must have a high accuracy against the actual topographic model. A model which can compute heights responding to random plane position by using of the topographic data and interpolation must be constructed. Interpolation affected by the accuraccy of the observations included noise, which affected by the slop and curvature weight. Data smoothing is a method to reduce the noise. Average declination and area ratio are variable which result similarity in according to slope. But in local area, area ratio well shows a local change. This study try to classify the terrain by the declination to analysis the effects of the declination and curvature weights, and then to represent the most probable model. The result are following : In terrain classification by the slop, p16 and p24 were fitted in the plane surface fit p16 and S in the varying surface, and S and p24 in the irregular surface in classification by curvature, p24 and S were fitted in the plane or varying surface, and p16 in the irregular surface In case of hybrid, p16, p24 and S are fitted in the plane, varying and irregular surface respectively. Smoothing is the most effective in case of slope of 50 persentage and of curvature weight of 0.0015.

• Journal of Life Science
• /
• v.29 no.7
• /
• pp.762-768
• /
• 2019

Doellingeria scabra Thunb. (syn. Aster scaber Thunb.), a perennial herb in the family Asteraceae, is frequently found in the wild mountain regions of Korea. This aim of this work was to measure the characteristics of patchiness of D. scabra in a local population on Mt. Maebong in Taeback-ci, Gangwon-do. The spatial distribution pattern of this species was estimated by analyzing ecological data by methods including the index of dispersion, Lloyd's mean crowding, and Morisita's index. The mean population density of the D. scabra population was 2.94. The D. scabra individuals were uniformly or randomly distributed in small-scale plots and were aggregately distributed in two large-scale plots ($16{\times}32m^2$ and $32{\times}32m^2$). The mean crowding ($M^*$) was 0.916. The mean patchiness index (PAI) was 0.796. Morisita's coefficient tended to decrease the density of the population as the plot size increased. The expected value of Eberhardt's index ($I_E$) in the local population was 2.623. Moran's I of D. scabra significantly differed from the expected value in 6 of 8 cases (75.0%). The first five classes were positive, with four showing statistical significance, indicating similarity among individuals in the first four distance classes (I-IV, 8 m), The results presented here could provide a theoretical basis for the conservation of D. scabra (Korean: chamchwi) and for the rehabilitation and sustainable management of forest ecosystems on Mt. Maebong, as well as on other mountains.

• The Korean Journal of Ecology
• /
• v.16 no.4
• /
• pp.515-526
• /
• 1993

In order to study the ecotypic variation of Rohinia pseudo-acacia L. distributed in southern area of Korean peninsula, 15 local populations(Daejin, Sokcho, Kangneung, Mt. Surak, Hongcheon, Kwangneung, Namhansanseong, Chungju, Yesan, Andong, Jeonju, Dalseong, Changweon, Mokpo and Wando), located from $34^{\circ}18'N\;to\;38^{\circ}36'N$, were selected based on the latitudes and geographical distances. Seeds of these populations were collected and protein contents of seeds and their band patterns were investigated. The seed proteins of all populations were electrophoresed on SDS-polyacrylamide gel. Total number of protein bands were 35, whose molecular weights ranged from 17, 258 daltons to 142, 232 daltons. The number of bands of seed proteins was 23 in Dalseong and Hongcheon and was 32 in Daejin and Sokcho, showing an increasing tendency in the number of bands as the latitude goes high. The local populations were classified into 3 local types based on protein analysis: the middle north east coastal type(Daejin, Sokcho. Kangneung), the central type (Mt. Surak, Hongcheon, Kwangneung, Namhansanseong, Chungju) and the southern type(Yesan, Andong, Jeonju, Dalseong, Changweon, Mokpo, Wando). According to the results of cluster analysis by UPGMA based on the similarity index(c0efficient of Jaccard) of the patterns, 3 local types were subdivided further into 6 types: the middle north east coastal type(Sokcho, Kangneung), the north central type I (Mt. Surak, Hongcheon), the north central type II (Narnhansanseong, Chungju, Daejin), the north central type III (Kwangneung), the south central type (Yesan, Dalseong, Jeonju) and the southern type(Andong, Changweon, Mokpo, Dalseong, Wando). The No. 12 band of the separated seed proteins showed the highest colored density in the preparations from all the populations. The No. 11~13 and No. 23~28 bands also showed high densities. As a whole, southern type populations (Changweon, Mokpo, Wando) showed high protein contents and high colored density. Total protein contents of the seeds in each population were variable from 9. 68mg / g (Mt. Surak) to 17.30mg/g (Jeonju), showing an increasing trends toward low latitudes.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.167-183
• /
• 2023

In the recent field of recommendation systems, various studies have been conducted to model sparse data effectively. Among these, GLocal-K(Global and Local Kernels for Recommender Systems) is a research endeavor combining global and local kernels to provide personalized recommendations by considering global data patterns and individual user characteristics. However, due to its utilization of kernel tricks, GLocal-K exhibits diminished performance on highly sparse data and struggles to offer recommendations for new users or items due to the absence of side information. In this paper, to address these limitations of GLocal-K, we propose the GEase-K (Global and EASE kernels for Recommender Systems) model, incorporating the EASE(Embarrassingly Shallow Autoencoders for Sparse Data) model and leveraging side information. Initially, we substitute EASE for the local kernel in GLocal-K to enhance recommendation performance on highly sparse data. EASE, functioning as a simple linear operational structure, is an autoencoder that performs highly on extremely sparse data through regularization and learning item similarity. Additionally, we utilize side information to alleviate the cold-start problem. We enhance the understanding of user-item similarities by employing a conditional autoencoder structure during the training process to incorporate side information. In conclusion, GEase-K demonstrates resilience in highly sparse data and cold-start situations by combining linear and nonlinear structures and utilizing side information. Experimental results show that GEase-K outperforms GLocal-K based on the RMSE and MAE metrics on the highly sparse GoodReads and ModCloth datasets. Furthermore, in cold-start experiments divided into four groups using the GoodReads and ModCloth datasets, GEase-K denotes superior performance compared to GLocal-K.

• (The) Research of the performance art and culture
• /
• no.19
• /
• pp.88-123
• /
• 2009

Gyeongsanjain Dano Festival is a local festival of Yeongnam that was initially designated as Intangible Cultural Heritage No. 44 "Hanjanggun Nori" in 1971 and finally became Gyeongsanjain Dano Festival in 2007. At first, few parts of it were designated for preservation and it was succeeded as a whole without connections between each part. Problems also rise from the fact that it has not been closely studied. Gyeongsanjain Dano Festival is mostly classified into five parts: Keungut, Yeowonmu, Hojanggut, Hanmyoje, and Jainpalgwangdae. This study has closely discussed each part as follows: 1. Keungut 1) Process and Contents of Keungut - Characteristics of Gut in Daegu and Gyeongsan 2) Presence of Performers that can Execute Traditional Gut of Local Area 3) Problems with the Name of Keungut - Ex: 'Hanjanggungut' or 'Keungut' 4) Problems of Dano Festival without Parts to Summon and Send off Spirits before and After Keungut and Restoration Measures 2. Yeowonmu 1) Essence of Yeowonmu and Yeowonhwa 2) Problems with the Mass Game of Hundreds of High School Girls 3) Origin and Succession of Yeowonmu 3. Hojanggut 1) Changes in Characters of Hojanggut 2) Composition and Characteristics of Hojanggut 3) Problems with the Name and Characteristics of Hojanggut 4. Jainpalgwangdae Nori 1) Similarity with Newly Created Shows 2) Problems with the Name of Palgwangdae 3) Difference with the Composition of Other Mask Dances 4) Dances and Movements Distinguished from Other Mask Dances in Yeongnam The following are the solutions suggested for Gyeongsan Jain Dano Festival. First, for the restoration of Gyeongsanjain Dano Festival, Dano festivals and Byeolsinje of nearby areas with clear traditions could be benchmarked. Second, the major content of Gyeongsanjain Dano Festival is 'Hanjanggungut,' and it has to be the leading content of the festival. Third, the structural principle of Korean festival must be adopted and the process and principle must correspond to those of other traditional shows. Fourth, as Gyeongsanjain Dano Festival is the comprehensive form of art, folk, and festival, each part must be closely related in a well-planned scenario. Fifth, Intangible talents and successful training must be widely acknowledged for successful transmission and responsible performances. Sixth, traditional festivals must be restored along with the discovery and development of various festival repertories and spectacles such as tour programs and experiential tours that contribute to local economy.

• Korean Journal of Environment and Ecology
• /
• v.25 no.3
• /
• pp.327-357
• /
• 2011

This study compared the species composition and distribution of the forest communities between Korea and Japan using vegetation releve database. The study included the eastern and southern Korean Peninsula, Is. Jeju, Is. Ulreung, northern Kyushu, central part of Japan and Is. Tsushima and divided in fourteen region with major mountain as the center. Seventy-nine forest communities were classified by tabular comparison with the total of 1844 releves. In the composition table, fifty-four groups of plant species were listed. Some grew in one area exclusively, while others were present in multiple locations. This study showed the similarity and heterogeneity of species composition among study areas. We grouped fourteen types showing similar physiognomy for forest communities. The physiognomical features of the vertical vegetation zones in the study area were similar, however, dominant species and species composition were different between the regions. In the lowland zone, Castanopsis-Type dominated by Castanopsis cuspidata var. sieboldii or Castanopsis cuspidata were distributed in the study area except for northern part of the Korean Peninsula and Is. Ulreung. Additionally, Persea thunbergii-Type was distributed widely in the lowest part of the study area, however, it showed the simplification of the species composition in the Korean Peninsula and Is. Ulreung. In the hilly zone, evergreen forest composed by Abies firma-Type distributed in northern Kyushuand Is. Tsushima, but it was absent in the other study region. The difference in forest community was conspicuous, especially in the montane zone. Deciduous broadleaved forests composed Quercus(D)-Type and Fagus-Type widely distributed in the montane of study area. However, forest community dominated by Quercus mongolica and Fagus multinervis flourished in Korea. On the otherhand, forest community dominated by Quercus mongolica var. grosseserrata, Fagus crenata and F. japonica distributed in Japan. In the sub-alpine zone, evergreen coniferous forest composed Abies-Quercus(D)-Type, Abies koreana-Type, and Pinus pumila-Type were distributed in the Korean Peninsula and Is. Jeju. Forests composed Taxus cuspidata var. nana-Type, Abies mariesii-Type, and Pinus pumila-Type distributed in the central part of Japan. As a result of DCA ordination, evergreen broadleaved forests of the whole study area showed the lower scores along the first axis, while deciduous and coniferous forests showed the higher scores. The forest types of western part of Japan were located the higer scores, forest types of the Korean Peninsula were located the lower scores, and forest types of Is. Jeju located in the middle. Warmth index (WI) and Annual range of temperature in climate factor were highly correlated on the first axis. The first axis reflected the gradient from oceanic climatic to continental one. The higher the altitude and further geographically, the lower the similarity among communities and the peculiarity of community appeared stronger. The historical background reflecting local flora has strongly influenced on development of community.

• Journal of Aquaculture
• /
• v.13 no.2
• /
• pp.129-135
• /
• 2000

Genetic variabilities of local populations of Anthocidaris crassispina and Hemicentrotus pulcherrimus were analysed by random amplified polymorphic DNA (RAPD) technique with phenotypic variabilities in timing of gonadal maturation, reproductive output and size of individuals in Cheju. H, pulcherrimus, collected from 4 locations during March 1997, indicated that Daepo individuals were significantly smaller than that at Hamdok, Wimi site A and Wimi site B (ANOVA, p>0.001). Gonodal-somatic index (GSI) of the Wimi site B population was significantly higher than that of three other locations (ANOVA, P>0.0001). An ANOVA test conducted on test size of A. crassispina, harvested from six different locations of Cheju during June 1997, indicated that the size of individuals from Pophwan was significantly smaller (P>0.0001) than that from five other locations. GSIs of urchin m Wimi and Hanlim were significantly higher than that from Pohwan and Oedo (ANOVA, p>0.0001). Genetic similarity, calculated from k13 primer analysis of total DNA, among the six different populations of A. crassispina varied from 0.67 to 0.92, and the values for H. pulcherrimus from 0.60 to 0.73; thus there was no genetic variation among different populations of the same species. Therefore, the populations are genetically homologous and the observed phenotypic variabilities were possibly associated with water temperature and food.