• 제목/요약/키워드: Cluster Based

검색결과 4,029건 처리시간 0.033초

User-Perspective Issue Clustering Using Multi-Layered Two-Mode Network Analysis (다계층 이원 네트워크를 활용한 사용자 관점의 이슈 클러스터링)

  • Kim, Jieun;Kim, Namgyu;Cho, Yoonho
    • Journal of Intelligence and Information Systems
    • /
    • 제20권2호
    • /
    • pp.93-107
    • /
    • 2014
  • In this paper, we report what we have observed with regard to user-perspective issue clustering based on multi-layered two-mode network analysis. This work is significant in the context of data collection by companies about customer needs. Most companies have failed to uncover such needs for products or services properly in terms of demographic data such as age, income levels, and purchase history. Because of excessive reliance on limited internal data, most recommendation systems do not provide decision makers with appropriate business information for current business circumstances. However, part of the problem is the increasing regulation of personal data gathering and privacy. This makes demographic or transaction data collection more difficult, and is a significant hurdle for traditional recommendation approaches because these systems demand a great deal of personal data or transaction logs. Our motivation for presenting this paper to academia is our strong belief, and evidence, that most customers' requirements for products can be effectively and efficiently analyzed from unstructured textual data such as Internet news text. In order to derive users' requirements from textual data obtained online, the proposed approach in this paper attempts to construct double two-mode networks, such as a user-news network and news-issue network, and to integrate these into one quasi-network as the input for issue clustering. One of the contributions of this research is the development of a methodology utilizing enormous amounts of unstructured textual data for user-oriented issue clustering by leveraging existing text mining and social network analysis. In order to build multi-layered two-mode networks of news logs, we need some tools such as text mining and topic analysis. We used not only SAS Enterprise Miner 12.1, which provides a text miner module and cluster module for textual data analysis, but also NetMiner 4 for network visualization and analysis. Our approach for user-perspective issue clustering is composed of six main phases: crawling, topic analysis, access pattern analysis, network merging, network conversion, and clustering. In the first phase, we collect visit logs for news sites by crawler. After gathering unstructured news article data, the topic analysis phase extracts issues from each news article in order to build an article-news network. For simplicity, 100 topics are extracted from 13,652 articles. In the third phase, a user-article network is constructed with access patterns derived from web transaction logs. The double two-mode networks are then merged into a quasi-network of user-issue. Finally, in the user-oriented issue-clustering phase, we classify issues through structural equivalence, and compare these with the clustering results from statistical tools and network analysis. An experiment with a large dataset was performed to build a multi-layer two-mode network. After that, we compared the results of issue clustering from SAS with that of network analysis. The experimental dataset was from a web site ranking site, and the biggest portal site in Korea. The sample dataset contains 150 million transaction logs and 13,652 news articles of 5,000 panels over one year. User-article and article-issue networks are constructed and merged into a user-issue quasi-network using Netminer. Our issue-clustering results applied the Partitioning Around Medoids (PAM) algorithm and Multidimensional Scaling (MDS), and are consistent with the results from SAS clustering. In spite of extensive efforts to provide user information with recommendation systems, most projects are successful only when companies have sufficient data about users and transactions. Our proposed methodology, user-perspective issue clustering, can provide practical support to decision-making in companies because it enhances user-related data from unstructured textual data. To overcome the problem of insufficient data from traditional approaches, our methodology infers customers' real interests by utilizing web transaction logs. In addition, we suggest topic analysis and issue clustering as a practical means of issue identification.

A Study on the Use of Wintering Habitats of Water Birds Arriving at Coastal Wetlands in Jeollanam Province, Korea (전라남도 연안습지에 도래하는 수조류의 월동지 이용에 관한 연구)

  • Choi, Young-Bok;Jung, Sook-Hee;Yoo, Seung-Hwa;Kang, Tae-Han;Lee, Han-Soo;Paek, Woon-Kee;Choi, Chung-Gill;Kim, In-Kyu
    • Korean Journal of Environment and Ecology
    • /
    • 제21권3호
    • /
    • pp.197-206
    • /
    • 2007
  • This study was conducted to survey the population of water birds wintering at the seven coastal wetlands of Jeollanam province including Suncheon Bay and Yeongsan Lake, from 2000 through 2003. The 90 species and 857,570 individuals in total were sighted at the seven survey sites. We classified the wintering water birds into seventeen groups of taxa based on the similar ecological attributes, among which, eight groups were found to inhabit the water surface or riparian areas. Classified groups that showed higher rate of using bay areas than that of lake areas were in the order of waders, gulls and swans. On the other hand, the groups that showed higher rate of using lake areas than that of using bay areas were revealed in the order of dabbling ducks, grebes and geese. In conclusion, there was a difference in the pattern between the two classified groups. As a result of the UPGMA cluster analysis using CCs ($S{\varnothing}rensen'a$ index of similarity and Ro (Horn's index of community overlap), the results showed that Suncheon Bay had the most unique species formation out of the seven areas. Bay and lake areas were different from each other in the formation of species and Individuals. As a result of combining the index rank according to the maximum aggregate count, the Suncheon Bay is ranked the highest in importance of the habitats for water birds, followed by the order of Boseong-Deukryang Bay, Gangjin Bay, Gocehongam Lake, Geumho Lake, Yeomam Lake, and Yeongsan Lake. Considered overall, the importance of the bay areas was relatively higher than that of reclaimed lake areas.

Community Structure and Health Assessment of Macrobenthic Assemblages at Spring and Summer in Garorim Bay, West Coast of Korea (가로림만에 서식하는 대형저서동물의 춘계와 하계의 군집구조 및 건강도 평가)

  • Jung, Rae-Hong;Seo, In-Soo;Lee, Won-Chan;Kim, Hyung-Chul;Kim, Jeong-Bae;Choi, Byoung-Mi;Yun, Jae-Seong;Na, Jong-Hun
    • Journal of the Korean Society of Marine Environment & Safety
    • /
    • 제20권5호
    • /
    • pp.491-503
    • /
    • 2014
  • This study was performed to investigate the community structure and health assessment of macrobenthic assemblages in Garorim Bay, West Coast of Korea. Macrobenthos were collected by van Veen grab sampler at May(spring) and July(summer) 2012. A total of 247 species occurred and mean density was $1,625\;ind.\;m^{-2}$, both of which were dominated by annelid polychaetes(120 species and $1,241m^{-2}$). Dominant species were the polychaetes Ampharete arctica, Lumbrineris longifolia, Mediomastus californiensis and Euclymene oerstedi, with a density of 445(${\pm}1,837\;ind.\;m^{-2}$), 103(${\pm}148\;ind.\;m^{-2}$), 55(${\pm}83\;ind.\;m^{-2}$) and 50(${\pm}104\;ind.\;m^{-2}$), respectively. The study area was divided into 3 station group based on the cluster analysis and nMDS ordination. These assemblage were : 1)the group 1 and 2 were associated with coarse sediment dominated stations and 2)the group 3 was connected with a mixed and fine sediment dominated stations group. The BPI and AMBI index were applied to assess the benthic ecological status. The ecological status of the Garorim Bay was "good status(slightly polluted)" to "high status(normal)" at most sampling stations during spring and summer. In conclusion, the two marine biotic index calculated shown that the Garorim Bay had a good ecological status.

A Study on the Characteristics of Vegetation Landscape of Fortress of Jeonju District in Represented on the (<전주지도>에 표현된 조선 후기 전주부성의 식생경관상)

  • Kang, In-ae;Rho, Jae-hyun
    • Journal of the Korean Institute of Traditional Landscape Architecture
    • /
    • 제36권2호
    • /
    • pp.1-10
    • /
    • 2018
  • This study aims to find out the characteristics of the vegetation landscape characteristics and system which led the formation of the urban image in Jeonju in the late Joseon period connected with urban spatial structure, using designated as treasure No. 1586 which was made in the middle of 18C. The vegetation landscape characteristics of Jeonju in the late Joseon Dynasty derived from the analysis of are summarized as follows. Firstly, the vegetation landscape system in Jeonju is composed of the natural vegetation around mountain area of Jeonju-Buseong, the independent vegetation or cluster planting forests linked with the main facilities, the Bibo-Forests connected with topographical characteristics of Jeonju, and the vegetation combined with a private garden. Secondly, planting landscape was specialized using flag species and local species. Thirdly, the garden-type plantation centered on the back yard or front of main facilities, with the background of natural vegetation landscape combined with the mountain area and the vegetation combined with a private garden, dominates vegetation landscape of Jeonju Buseong as objects. Fourthly, in order to overcome the defects of topographical characteristics, the Bibo-Forests were emphasized as an important planting landscape element in addition to the vegetation landscape elements connected with main facilities. Fifth, ecological vegetation landscape technique was taken considering the topographical characteristics. The characteristics of vegetation landscape of Jeonju Buseong, which is derived from , have an important meaning to restore and reproduce Jeonju's historical features. Especially, the vegetation communities of the non-booming concept combined with the geographical features, the ecological landscape harmonizing with the topography, the round house type landscape mixed with the private house, and the specialization of vegetation landscape using local species are important factors in securing the city image based on the historical characteristics and creating a city brand that utilizes vegetation landscape.

Macrobenthic Community at the Subtidal Area Around Taebudo in Kyeonggi Bay, Korea (경기만 대부도 주변 조하대 해역의 저서동물 군집)

  • LIM Hyun-Sig;CHOI Jin-Woo
    • Korean Journal of Fisheries and Aquatic Sciences
    • /
    • 제31권4호
    • /
    • pp.453-462
    • /
    • 1998
  • Macrobenthic community structure was studied at thirteen stations in Taebudo subtidal area, Korea, from July to October 1996. Triple macrobenthos samples were collected using a van Veen grab (0.1 $m^2$) at each station during the study period. A total of 209 species of macrobenthos was sampled with a mean density of 1,093 ind./$m^2$ and biomass of 134,86 g/$m^2$. Of these, there were 72 species of polychaetes ($34.5\%$), 69 crustaceans ($33.0\%$) and 49 molluscs ($23.4\%$). Polyalaetes were represented as a density-dominant faunal group with a mean density of 608 ind./$m^2$, comprising $55.6\%$ of the total benthic animals. It was followed by crustaceans with 307 ind./$m^2$($28.1\%$ of the total density), Echinoderms were represented as a biomass-dominant faunal group with a mean biomass of 54.21 g/$m^2$($40.2\%$ of total biomass). Total number of species and diversity were low in the inner part of the study area with high mud content and high in the offshore stations of mixed sediments. Major dominant species were three polychaetes, Heteromastus filifomis, Scoloplos armiger and Tharyx sp. whose mean densities were 70 ind./$m^2$, 67 ind./$m^2$, and 66 ind./$m^2$, respectively. Cluster analysis showed that the study area could be divided into five stational groups based on the faunal composition, that is, the innermost stations, coastal stations, transitional stations and two offshore station groups. The species diversity of these groups increased from the inner stational group toward the outer groups.

  • PDF

Growth Environment and Vegetation Structure of Cephalotaxus koreana Nakai in South Korea Natural Habitats (국내 개비자나무 자생지 생육환경 및 식생구조)

  • Kim, Young Ki;Kim, Joon Seon;Lee, Kap Yeon;Kim, Moon Sup
    • Korean Journal of Plant Resources
    • /
    • 제31권4호
    • /
    • pp.384-395
    • /
    • 2018
  • This study was carried out to investigate the environment factors including community structure and soil characteristics in the wild habitats of Cephalotaxus koreana, and offers the basic information for habitats conservation and restoration. Most of the wild habitats were located at altitudes between 148~835 m with inclinations ranged as $12{\sim}32^{\circ}$. The average soil pH was 4.7~5.9, soil organic matter was 5.72~15.99%, cation exchange capacity was $14.1{\sim}19.9cmolc/kg^{-1}$ and exchangeable $K^+$, $Ca^{2+}$, $Mg^{2+}$ was 0.25~0.48 cmolc/kg, 0.79~6.68 cmolc/kg, 0.31~1.73 cmolc/kg, respectively. The dominant species of tree layer were found to be dominated by Quercus dentata in Jekbo-san (C1), Acer pictum in Bogae-san (C2), Acer pseudosieboldianum in Geumwon-san (C3), Q. serrata in Jiri-san (C4), Zelkova serrata in Baegun-san (C5), and Q. acutissima in Duryun-san (C6). The Species diversity (H') was 0.854~1.234, evenness (J') was 0.654~0.993, and dominance (D) was found to be 0.067~0.346. Correlation coefficients analysis based on environmental factors, community structure and value of species diversity shows that growth of Cephalotaxus koreana is correlated with species diversity and evenness. This result show that Cephalotaxus koreana habitats located in mature stands.

Analysis for the Major Traits and Genetic Similarity of Native Ginseng (Panax Ginseng C.A. Meyer) Collections in Korea. (인삼(Panax ginsneg C.A. Meyer) 수집종의 주요 특성 및 유연관계 분석)

  • Rhim, Soon-Young;Sohn, Jae-Keun;Ryu, Tae-Seok;Kwon, Tae-Ryong;Choi, Jin-Kook;Choi, Hong-Jib
    • Korean Journal of Breeding Science
    • /
    • 제42권5호
    • /
    • pp.488-494
    • /
    • 2010
  • In this study, the major agronomic traits were investigated and RAPD technique was applied for the analysis of the genetic relations between the native ginsengs collected from Poonggi and Geumsan provinces in Korea. The main morphological traits were measured for a total of 54 collections of native ginseng from two areas based on UPOV standard. A total of 58 collections consisting of twenty-one native ginsengs collections from Poonggi area, twenty-nine collections from Geumsan area and four varieties of P. quinquefolium, P. japonicum, Chunpoong and Hwangsuk as controls were analyzed and clustered by RAPD. The results indicated that 01-9, 01-35 and 01-44 collections from Poonggi area were grouped into Geumsan area, while 332001, 332002 and 332003 collections from Geumsan area were grouped into Poonggi area. On comparison to the similarity of Poonggi collections (73-95%), the Geumsan collections showed 65-86% similarity in the population. Thus, the cluster should be applied according to the number of stem, number of leaves per stem and leaflet shape on the regionally native ginseng collections. The fourteen primers such as OPA02, OPA07, OPC08, OPD11, OPD20 and so on, will be used to select the native ginseng in the future studies.

Characteristics of Panicle Traits for 178 Rice Varieties Bred in Korea (국내에서 육성된 벼 품종들의 이삭형질 특성)

  • Park, Hyun-Su;Kim, Ki-Young;Mo, Young-Jun;Choung, Jin-Il;Kang, Hyun-Jung;Kim, Bo-Kyung;Shin, Mun-Sik;Ko, Jae-Kwon;Kim, Sun-Hyung;Lee, Bu-Young
    • Korean Journal of Breeding Science
    • /
    • 제42권2호
    • /
    • pp.169-180
    • /
    • 2010
  • This study was conducted to investigate characteristics of panicle traits which are important factors affecting yield and grain quality of rice. Twelve panicle traits in 178 Korean rice varieties composed of 160 Japonica type varieties and 18 Tongil type varieties were investigated. Tongil type varieties had longer panicle and thicker neck node than Japonica type varieties. Other traits such as number of total spikelets, total rachis-branches, secondary rachis-branches (SRBs) per panicle, total spikelets on SRBs per panicle, mean number of spikelets on a SRB, and mean number of SRBs per primary rachis branch (PRB) in Tongil type varieties were also higher than in Japonica type varieties. On the other hand, Japonica type varieties were shown to have well exserted panicle and little more mean number of spikelets on a PRB than Tongil type varieties. According to cluster analysis based on 12 panicle traits, 178 varieties were divided into four main groups. Group I had 133 Japonica type varieties and was characterized by relatively well exserted short panicle, small thickness of neck node, few rachis-branches and little sink size than other group. Group II was composed of 24 Japonica type varieties and 6 Tongil type varieties showing medium value and range between Group I and III. Group III included 11 Tongil type varieties and 1 Japonica type variety 'Baegjinju1' characterized by relatively poor exserted long panicle, big thickness of neck node, many rachis-branches and large sink size. Group IV was solely composed of 'Nongan', which had well exserted long panicle, big thickness of neck node, many rachis-branches and large-sink size. In correlation analysis, number of total spikelets per panicle showed very high correlation with the number of total rachis-branches per panicle (r=0.975), number of spikelets on SRBs per panicle (0.962), number of SRBs per panicle (0.959), mean number of SRBs per PRB (0.746) and mean number of spikelets on SRBs (0.738).

Changes in Aquatic Insect Community Structure in Wonju Stream based on a Comparison of Previous Studies (과거 문헌 비교를 통한 원주천 수서곤충 군집구조 변화)

  • Han, Jung Soo;Choi, Jun Kil;Won, Kyung Ho;Lee, Hwang Goo
    • Korean Journal of Environmental Biology
    • /
    • 제36권3호
    • /
    • pp.400-411
    • /
    • 2018
  • This study was a survey of the Wonju stream in Wonju city from May 2015 to September 2016. A total of three sites were selected from the upstream area Gwanseol-dong to the downstream area Hojeo-myeon. Physicochemical analysis, aquatic insect changes, cluster analysis, functional group analysis, rarefaction curve, and statistical analysis were compared between 2004 and 2016. A total of 19 species (38.78%) in 2004 and 22 species (36.67%) in 2016 were analyzed, with the largest number belonging to ephemeroptera. The individual ratio ranged from 27,759.2 (ind. $m^{-2}$, 84.30%) in 2004 to 4,573.2 (ind. $m^{-2}$, 41.64%) in 2016, with the highest number involving diptera. As a result of the community analysis, significant differences were detected in the indices of dominance, diversity, evenness, and richness in 2004 and 2016 (p<0.05). Burrowers of the habitat orientation groups showed the greatest variation with an average of -68.00% (${\pm}2.15$) and the collector-gatherers of the functional feeding groups showed the highest variation of -40.12% (${\pm}1.77$). The rarefaction curve analysis suggested that the species was the poorest in the midstream regions in 2004 and 2016. Physical factors and water quality showed a significant correlation with diversity index, evenness index, and the number of individuals. MDS analysis of the similarity of upstream and downstream regions was high in 2004, and low in 2016. The differences were attributed to physicochemical changes such as increase in flow velocity due to improvement of small dams and changes in bottom structure.

Metallurgical Study on the Iron Artifacts Excavated from Sudang-ri Site in Geumsan (금산 수당리유적 출토 철제유물의 금속학적 연구)

  • Park, Hyung-ho;Cho, Nam-chul;Lee, Hun
    • Korean Journal of Heritage: History & Science
    • /
    • 제46권3호
    • /
    • pp.134-149
    • /
    • 2013
  • The Sudang-ri Site in Geumsan is considered the historic site where Baekje dominated the inland traffic route to Gaya through Geumsan and Jinan in the 5th Century. This study identified the production techniques of iron by conducting an analysis of metallographical microstructure of the artifacts such as an iron sword and an iron sickle that were excavated in Sudang-ri Site, Geumsan, one of the regions ruled by Baekje, and tried to figure out the characteristics and the technical systems of Baekje's ironmaking around the 5th Century by comparing them with other iron artifacts produced around the same time. The analysis showed that various production techniques were applied to the artifacts excavated in Sudang-ri Site, Geumsan. Depending on the production techniques, they can be divided largely into three methods: the simple shape-forging method, the steel manufacture method after forging, and the steel manufacture & heat-treatment method after forging. The iron sickle from the stone chamber tomb No. 1, which was produced only through forging, is mostly composed of soft ferrite at both edges of the blade and at the rear making the use of the weapon impractical. From this fact, it is presumed that they were produced as burial objects or ceremonial accessories for the person buried. The iron axe from the outer stone coffin tomb No. 1 and the iron swords and sickle from the outer stone coffin tomb No. 12, which were produced through the steel manufacture method after forging such as carburizing, did not go through the heat treatment such as quenching, but applied different production processes to each part. Therefore, it is deemed that they were produced as daily tools for cultivation rather than burial objects or ceremonial accessories. The production techniques following the forging process - carburizing and heat treatment - can be found on the iron swords from the outer stone coffin tomb No. 5 and the outer stone coffin tomb No. 12. The sturdy structure of the blade part and the durable structure of the rear processed with heat are deemed to have been produced as weaponry and used by the person buried. Based on the analysis of the iron artifacts excavated from Sudang-ri Site in Geumsan, the characteristics of iron production techniques were investigated by comparing them with the artifacts from Yongwon-ri Site in Cheonan, Bongseon-ri Site in Seocheon, and Bujang-ri Site in Seosan that were made around the same time as the cluster of Baekje tombs examined by the metallographical microstructure analysis of this study. For the iron artifacts analyzed here, the changes in the techniques were investigated using the iron swords common in all of the tombs. In the case of the iron swords, it was identified the heat treatment technique called tempering was applied from the 4th Century.