• Journal of Korean Society of Forest Science
• /
• v.107 no.3
• /
• pp.245-257
• /
• 2018

The study was carried out to discover the forest vegetation structure in Hongdo Island, Jeonnam province. Vegetation data were collected by total of forty one quadrate plots using Z-M phytosociological method from June to August in 2017, and analyzed by vegetation classification, mean importance value and species diversity. As a result of vegetation type classification, Castanopsis sieboldii community group was classified at a top level of vegetation hierarchy. In the level of community, it was classified into Neolitsea sericea community and Carpinus turczaninowii community. N. sericea community was subdivided into Ficus erecta group(Vegetation unit 1) and Arisaema ringens group(VU 2). C. turczaninowii community was subdivided into Fraxinus sieboldiana group(VU 3) and C. turczaninowii typical group(VU 4). Therefore, it was classified into total of four vegetation units(one community group, three communities and four groups). As a result of mean importance value, Castanopsis sieboldii was the highest in VU 1, VU 2, VU 4, and C. turczaninowii in VU 4, respectively. In case of species diversity, VU 3 showed the highest among four units in species diversity index. In conclusion, the forest vegetation of Hongdo Island was classified into four units and seven species groups. Hongdo Island could be conclusively managed by community ecological approach for the units and groups. Also it was considered that a research for the succession to the evergreen broad-leaved forest should be more intensively proceeded near future.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.6 no.2
• /
• pp.42-52
• /
• 1988

The main object of this study is to prove the effectiveness of Landsat data in detecting the stressed areas in forest by extracting these areas. And also to choose the effective bands for this type of survey and to reduce the effect of shadow in forest to improve the accuracy of classification are the other objects. In this study Landsat-5 TM data is used and image processing techniques such as spatial filtering and ratio are taken to reduce the effect of shadow and to improve the classification accuracy. As a result following conclusions are obtained. First, Landsat TM data is useful to detect the stressed areas in forest. Second, when detecting the stressed area, band 4 and 5 are the most effective. Third, spatial filtering and ratio are useful to reudce the effect of shadow and improve the classification accuracy. Especially, ratio has great effect on improving the classification accuracy between forest and other areas.

• Journal of the Korean Wood Science and Technology
• /
• v.49 no.5
• /
• pp.430-452
• /
• 2021

The recent social atmosphere has been a preference for wood utilization and woodworks. The general public does not have many opportunities to enjoy wood culture, so there is a lack of awareness and foundation of wood culture. In this study, classification and case analysis of wood culture were conducted as basic research for establishing a promotion strategy for the general public to enjoy wood culture. The specificity of wood culture and cultural resources was analyzed to establish the concept of wood cultural resources. Through the analysis, wood cultural resources were defined as products created as a result of human activities that implied the cultural value of wood and wood use in terms of conservation, discovery, and utilization. The types of wood cultural resources were classified into seven categories using the classification examples performed on cultural resources: cultural heritage, cultural facilities, cultural festival, wood architecture, culture contents, culture education, and wood products. In addition, cases were searched and proposed for each type of wood cultural resources.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.17 no.3
• /
• pp.35-51
• /
• 2014

A phytosociological vegetation survey was conducted in July to September 2011 in order to examine the vegetation community structure in Mt. Ohseo area. It was aimed to provide basic data for the effective vegetation conservation by analyzing the importance, species diversity and community similarity of the forest community in Mt. Ohseo for each layer, followed by the classification of the actual forest vegetation. According to the cluster analysis, the community type of Mt. Ohseo was classified into a total of 4 vegetation communities: Pinus densiflora community, Cornus controversa-Quercus serrata community, Miscanthus sinensis community, and Quercus mongolica community; the vegetation type 4 showed the lowest species diversity index of 0.5236, and vegetation type-2 showed the highest species diversity index of 0.6606. The community similarity between Quercus mongolica community and Pinus densiflora community showed the highest 0.679, and the community similarity between Quercus serrata community and Pinus densiflora community and between Quercus serrata community and Quercus mongolica community showed the levels of 0.5, respectively.

• Proceedings of the KSRS Conference
• /
• 2008.10a
• /
• pp.29-32
• /
• 2008

In this paper, we propose one unsupervised classification technique using the degree of polarization (DoP) and the co-polarized phase-difference (CPD) statistics, instead of the entropy and alpha. It is shown that the DoP is closely related to the entropy, and the CPD to the alpha. The DoP explains the feature how much the effect of multiple reflections is contained. Hence, the DoP could be used as an important factor for classifying classes. The CPD can also be computed from the measured Mueller matrix elements. For the smooth surface scattering, the CPD is about $0^{\circ}$, and for dihedral-type scattering, the CPD is about $180^{\circ}$. A DoP-CPD diagram with appropriate boundaries between six different classes is developed based on the SAR image. The classification results are compared with the existing Entropy-alpha diagram as well as the IPL-AirSAR polarimetric data. The technique may have capability to classify an SAR image into six major classes; a bare surface, a village, a crown-layer short vegetation canopy, a trunk-layer short vegetation canopy, a crown-layer forest, and a trunk-dominated forest. Based on the DoP and CPD analysis, a simple three-component decomposition technique was also proposed.

• Journal of Information Technology Services
• /
• v.14 no.4
• /
• pp.159-169
• /
• 2015

Customer reviews are one of the important sources for purchase decision makings in online stores. Online stores have tried to provide useful reviews in product pages to customers. To assess the usefulness of customer reviews before other users have voted enough on the reviews, diverse aspects of reviews were utilized in prevous studies. Style and semantic information were utilized in many studies. This study aims to test diverse alogrithms and datasets for identifying a proper classification method and threshold to classify useful reviews. In particular, most researches utilized ratio type helpfulness index as Amazon.com used. However, there is another type of usefulness index utilized in TripAdviser.com or Yelp.com, count type helpfulness index. There was no proper threshold to classify useful reviews yet for count type helpfulness index. This study used reivews and their usefulness votes on restaurnats from Yelp.com to devise diverse datasets and applied text mining approaches to classify useful reviews. Random Forest, SVM, and GLMNET showed the greater values of accuracy than other approaches.

• Journal of Environmental Science International
• /
• v.12 no.2
• /
• pp.93-100
• /
• 2003

In most mountainous parts of the temperate zone of Japan along the Pacific Ocean, some climatic climax forests, whose main dominant species is Fagus crenate, F. japonica or Quercus mongolica var. grosseserrata, are distributed. In the riparian regions of the zone, however, there appear summer green forests composed of the different species from the climatic climax forests. Climate plays an important role in determining the overall distribution of vegetation, but some environmental factors, i.e., topography, soil type, soil moisture content, etc. have a great influence on vegetation formation. Riparian forests seem to be controlled by various geomorphologic disturbances, such as landslide, soil erosion and accumulation. The study aims to present the relationships among vegetation, soils and landforms in the process of determining riparian forests dominated by Fraxinus platypoda and Pterocarya rhoifolia establishment in the mountainous region of central Japan. The study area extends an area of 302 ha with a range of elevation between 925 m and 1,681 m at the Chichibu mountains. The landforms were corditied at sampling grids (25 $\times$ 25 m, n = 4,843) using a hierarchical system, and a brief description of the forest soil classification was also given. The mutual relationship analysis indicated that forest soils and landforms play a significant role in determining the geomorphological process of riparian forest, and shaping the ultimate pattern of vegetation. At the study area, riparian forests were mainly found on the $B_E$ forest soil type and steep slopes ( > 30$^{\circ}$) at convex slopes along the streams. On the other hand, the direction of slopes did not have a significant impact on the establishment of the riparian forests. A mosaic of patchy distribution of those riparian forests on the slightly wetter $B_E$ forest soil type was one of the characteristic features of the study area. This particular soil which contained large talus gravels was found on the land formed by erosion and deposition of landslide.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.8 no.3
• /
• pp.132-140
• /
• 2006

Forest vegetation classification and species composition of Mt. Ilwol, Yeongyang-Gun, Korea, were studied combining the Braun-Blanquet approach with numerical syntaxonomical analyses (TWINSPAN). Vegetation types and various ecological characteristics such as flora, constancy classes, species ratio of life-form, species diversity and importance value were analyzed. Sixty-eight samples were taken from a $100m^2$ square plot each. Forest communities were identified as two great types: arid landform of mountainside (AM) and humid fertility of piedmont and valley (HP). The former was divided into 3 communities (Rhododendron mucronulatum, Quercus variabilis, Hosta capitat community) and 2groups, and the latter into 3 communities (Tilia amurensis, Vitis coignetiae, Philadelphus schrenckii community) and 2 groups. Vegetation was classified into 8 units. Floristically, the most represented family was Compositae with 26 species. Species with percentage constance degree of more than 61% was Quercus mongolica (72.1%, IV); Carex siderosticat (III) and Fraxinus rhynchophylla (III) were 50.0 and 41.1%, respectively. Life-forms species ratios for trees, subtrees, shrub, vines, grominoids, forbs and ferns were 18.5, 5.7, 14.9, 6.6, 8.8, 42.4 and 3.1%, respectively, PH type showed from $1.70{\pm}0.50\;to\;1.97{\pm}0.57$ and AM type was from $1.40{\pm}0.18\;to\;1.62{\pm}0.20$ in species diversity; therefore, the former type showed higher species diversity than the latter, According to importance value analysis, Pinus densiflora, Quercus mongolica and Q. variabilis were higher in the tree layer, Q. mongolica in the subtree layer, Fraxinus sieboldiana, R. schlippenbachii, etc. in the shrub layer and Carex siderosticta, Carex humilis, etc. in the herb layer.

• Journal of Korean Society of Forest Science
• /
• v.86 no.3
• /
• pp.365-377
• /
• 1997

Much of the Cumberland Plateau of Tennessee, USA is in mixed hardwoods for which there are no applicable growth and yield predictors. Use of site index as a variable in growth and yield prediction models is limited in most stands because their history is not known and many may not be even-aged. Landtypes may offer an alternative to site index for these mixed stands because they were designed to include land of about equal productivity. To determine vegetation by landtype, dependency between landtype and detailed forest type was tested with Chi-square. Differences in productivity among landtypes were tested by employing regression analyses and analysis of variance(ANOVA). Basal area growth was fitted to the nonlinear models developed by Moser and Hall(1969). Basal area growth and volume growth were also predicted as a function of initial total basal area and initial volume with linear regression by landtype and by landtype class. Differences in basal area growth and volume growth by landtype were tested with ANOVA. Dependency between site class and landtype was tested with Chi-square. Vegetation types seem to be related to landtypes in the study area although the validity of the test is questionable because of a high proportion of sparsely occupied cells. No statistically significant differences in productivity among landtypes were found in this study.

• Korean Journal of Environment and Ecology
• /
• v.15 no.3
• /
• pp.237-246
• /
• 2001

To investigate the forest structure and to suggest the basic data of forest in the Dongwol Valley, the Gyeryongsan National Park, forty one plots were set up and surveyed. According to the analysis of classification by TWINSPAN, the community was divided by seven groups of Carpinus taxinora(I), Styrax japonica(II), Pinus densiflora A type(III), Quercus acutissima(IV), P. densinora-Q. valiabilis(V), p. densinora B type(Ⅵ) and Q. mongolica community(Ⅶ). The structure of communities were investigated using importance value. The survey results were summarized as follows: 1) the results of annual ring analysis revealed that the age of forest in the Dongwol Valley was about 30~50 years old,2) number of the average species was 15.6$\pm$4.7 and number of average individual was 124.9$\pm$36.4 per a plot(100$\textrm{cm}^2$). From the above results, it was anticipated that C. laxiflora will be climax species instead of Q. spp. and P. densinora.