• Title/Summary/Keyword: multi-level categories

Search Result 55, Processing Time 0.031 seconds

HKIB-20000 & HKIB-40075: Hangul Benchmark Collections for Text Categorization Research

  • Kim, Jin-Suk;Choe, Ho-Seop;You, Beom-Jong;Seo, Jeong-Hyun;Lee, Suk-Hoon;Ra, Dong-Yul
    • Journal of Computing Science and Engineering
    • /
    • v.3 no.3
    • /
    • pp.165-180
    • /
    • 2009
  • The HKIB, or Hankookilbo, test collections are two archives of Korean newswire stories manually categorized with semi-hierarchical or hierarchical category taxonomies. The base newswire stories were made available by the Hankook Ilbo (The Korea Daily) for research purposes. At first, Chungnam National University and KISTI collaborated to manually tag 40,075 news stories with categories by semi-hierarchical and balanced three-level classification scheme, where each news story has only one level-3 category (single-labeling). We refer to this original data set as HKIB-40075 test collection. And then Yonsei University and KISTI collaborated to select 20,000 newswire stories from the HKIB-40075 test collection, to rearrange the classification scheme to be fully hierarchical but unbalanced, and to assign one or more categories to each news story (multi-labeling). We refer to this modified data set as HKIB-20000 test collection. We benchmark a k-NN categorization algorithm both on HKIB-20000 and on HKIB-40075, illustrating properties of the collections, providing baseline results for future studies, and suggesting new directions for further research on Korean text categorization problem.

Crowd Density Estimation with Multi-class Adaboost in elevator (다중 클래스 아다부스트를 이용한 엘리베이터 내 군집 밀도 추정)

  • Kim, Dae-Hun;Lee, Young-Hyun;Ku, Bon-Hwa;Ko, Han-Seok
    • Journal of the Korea Society of Computer and Information
    • /
    • v.17 no.7
    • /
    • pp.45-52
    • /
    • 2012
  • In this paper, an crowd density in elevator estimation method based on multi-class Adaboost classifier is proposed. The SOM (Self-Organizing Map) based conventional methods have shown insufficient performance in practical scenarios and have weakness for low reproducibility. The proposed method estimates the crowd density using multi-class Adaboost classifier with texture features, namely, GLDM(Grey-Level Dependency Matrix) or GGDM(Grey-Gradient Dependency Matrix). In order to classify into multi-label, weak classifier which have better performance is generated by modifying a weight update equation of general Adaboost algorithm. The crowd density is classified into four categories depending on the number of persons in the crowd, which can be 0 person, 1-2 people, 3-4 people, and 5 or more people. The experimental results under indoor environment show the proposed method improves detection rate by about 20% compared to that of the conventional method.

Verification of Logic Gate Interconnection (논리회로 상호간의 연결도 검증)

  • Jung, Ja Choon;Kyung, Chong Min
    • Journal of the Korean Institute of Telematics and Electronics
    • /
    • v.24 no.2
    • /
    • pp.338-346
    • /
    • 1987
  • This paper describes a method for verifying whether a given geometrical layout correcdtly reflects the original logic level description. The logic description extracted from layout data was directly compadred with the original logic diagram generated at logic level design stage where the logic diagram is represented as a weighted multi-place graph. The comparison is based on graph isomorphism and error messages(error categories and locations)are invoked if any difference is found between the two logic descriptions. An efficient partitioning algorithm which consists of two steps, candidate selection and equal weight partitioning procedure, enables the entire verification process to occur in O(n log n) time.

  • PDF

Medical Experiences and Unmet Health Care Perception among Elderly People with Chronic Disease (만성질환을 가진 노인의 의료경험과 미충족의료 인식)

  • Min, Dong-Hoo;Cho, Jung-Yeon;Kim, Jeong-Gil;Seo, Su-Jin;Kim, Mi-Kyung;Shim, Eun-Hye;Cha, Yu-Hyun;Kim, Chang-Yup
    • Health Policy and Management
    • /
    • v.28 no.1
    • /
    • pp.35-47
    • /
    • 2018
  • Background: The purpose of this study is to elucidate the context of medical experience and the perception of unmet healthcare of elderly people with chronic diseases based on in-depth interview data. Methods: We carried out in-depth interviews with 10 elderly people with chronic diseases using semi-structured questionnaires based on literature review. The in-depth interview data were analyzed using thematic analysis; one qualitative research methodology, three core meaning categories, and four attributes associated with unmet healthcare were ultimately derived. Results: The context of the medical experience were based on the following three categories: (1) discomfort due to diseases and high medical needs, (2) the poor community medical environment and difficulties in accessing to metropolitan medical institutions, and (3) inconvenience caused by long waiting time and side effects of medicine. In addition, the elderly with chronic disease realized the unmet healthcare as (1) the availability related to the desired medical institutions at the right time, (2) the affordability related to their economic capacity, (3) the effectiveness of the medical services they experienced, and (4) the appropriateness related to receiving medical services in a pleasant environment. Conclusion: The perception of unmet healthcare among the elderly with chronic disease is the result of interaction of multi-level and multi-dimensional factors related with their medical experience.

Workplace bullying among Korean registered nurses: A meta-aggregation of qualitative studies (국내 간호사의 직장 내 괴롭힘 경험: 질적 메타 종합연구)

  • Park, Eun-Jun;Kang, Hyunwook;Ko, Ji Woon
    • The Journal of Korean Academic Society of Nursing Education
    • /
    • v.29 no.4
    • /
    • pp.450-468
    • /
    • 2023
  • Purpose: This study aimed to conduct a qualitative synthesis of workplace bullying experiences among nurses in Republic of Korea. Methods: Following the PRISMA guideline, a literature search was conducted using seven domestic and three international databases. Studies published in Korean or English from inception to December 31, 2022 were included. A meta-aggregation approach suggested by the Joanna Briggs Institute was used to synthesize the research findings. Results: Fourteen studies were included in this review. As a result of a data analysis of the selected studies, 199 subthemes and supporting illustrations were identified and grouped into 36 related categories. Based on the subthemes and categories, five synthesized findings were developed: (1) the individual and organizational causes of workplace bullying; (2) the various types of physical violence and psychological harassment; (3) the negative impact of workplace bullying and its effect on self-growth; (4) active and passive coping efforts in dealing with bullying; and (5) strategies for preventing bullying incidents. Conclusion: Based on the synthesized findings, four recommendations were made: (1) improving the challenging working conditions for nurses; (2) enhancing educational programs for new nursing graduates; and (3) promoting proactive responses from nursing managers in conjunction with an expansion of resilience training for nursing students. Finally, to address the issue of workplace bullying, (4) multi-center and multi-level research involving nursing organizations needs to be conducted.

Business Intelligence Design for Strategic Decision Making for Small and Midium-size E-Commerce Sellers: Focusing on Promotion Strategy (중소 전자상거래 판매상의 전략적 의사결정을 위한 비즈니스 인텔리전스 설계: 프로모션 전략을 중심으로)

  • Seung-Joo Lee;Young-Hyun Lee;Jin-Hyun Lee;Kang-Hyun Lee;Kwang-Sup Shin
    • The Journal of Bigdata
    • /
    • v.8 no.2
    • /
    • pp.201-222
    • /
    • 2023
  • As the e-Commerce gets increased based on the platform, a lot of small and medium sized sellers have tried to develop the more effective strategies to maximize the profit. In order to increase the profitability, it is quite important to make the strategic decisions based on the range of promotion, discount rate and categories of products. This research aims to develop the business intelligence application which can help sellers of e-Commerce platform make better decisions. To decide whether or not to promote, it is needed to predict the level of increase in sales after promotion. I n this research, we have applied the various machine learning algorithm such as MLP(Multi Layer Perceptron), Gradient Boosting Regression, Random Forest, and Linear Regression. Because of the complexity of data structure and distinctive characteristics of product categories, Random Forest and MLP showed the best performance. It seems possible to apply the proposed approach in this research in support the small and medium sized sellers to react on the market changes and to make the reasonable decisions based on the data, not their own experience.

Developing System and Site Level Framework of Management Effectiveness Evaluation for the Forest Genetic Resources Reserve in Korea (산림유전자원보호구역의 관리효과성 평가를 위한 시스템 및 현장 수준의 평가틀 개발)

  • Lee, Dong-Ho;Kang, Mihee;Kim, Seong-il
    • Journal of Korean Society of Forest Science
    • /
    • v.105 no.4
    • /
    • pp.472-485
    • /
    • 2016
  • The main purpose of this research was to develop a multi-level evaluation framework for the management effectiveness of the Forest Genetic Resources Reserve (FGRR) at both the system level and the site level. The initial system level Management Effectiveness Evaluation (MEE) framework for FGRR was developed based on the MEE Framework designed by IUCN WCPA and MEE framework for Korean National Parks that was designed jointly by IUCN, the Korean Ministry of Environment, and the Korea National Park Service. Several indicators were added or modified considering characteristics of the FGRR. The final system level MEE frameworks consisted of 6 categories with total of 39 criteria and 42 indicators based on expert survey results. The initial site-level MEE framework was developed based on the site level MEE framework for Korean National Parks that was designed jointly by IUCN, the Korean Ministry of Environment, and the Korea National Park Service. The final site level MEE framework consisted of 6 categories with total of 16 criteria and 40 indicators based on both an expert survey and an intensive workshop with the officers in charge of managing the FGRR from the Korea Forest Service and local governments.

Construction of BIBFRAME-Based Linking Structure for Interrelating Bibliographic Records for Alternative Format Materials (대체자료 서지레코드 연결을 위한 BIBFRAME 기반 연계 구조 구축)

  • Lee, Seungmin
    • Journal of the Korean Society for Library and Information Science
    • /
    • v.55 no.3
    • /
    • pp.281-301
    • /
    • 2021
  • Bibliographic records for alternative format materials using a flat bibliographic structure based on unit record have limitations in that they cannot be linked with other related resources in record level, which makes it difficult to secure sufficient access to alternative format materials for people with disabilities. In order to address this problem, this research constructed a linking structure which adopted FRBR-based BIBFRAME structure to establish a bibliographic structure that can interrelate the bibliographic records with flat structure and the multi-level BIBFRAME structure. This research proposed a linking bibliographic structure that enables interlinking between bibliographic records for related alternative format materials by composing upper-level linking property categories and interconnecting them to classes and properties of BIBFRAME.

Fire Response Education for Hospital Healthcare Providers: A Scoping Review (병원 의료종사자 대상 화재 대응 교육 현황: 주제범위 문헌고찰)

  • Min-Ji Kim;Seung-Eun Lee;Hyun-Eun Park
    • Quality Improvement in Health Care
    • /
    • v.29 no.2
    • /
    • pp.32-46
    • /
    • 2023
  • Purpose: Fire response education is critical for healthcare providers working in hospitals to ensure a safe environment for patients and staff. However, a comprehensive review that thoroughly examines the contents, methodologies, and outcomes of fire response education in hospitals is currently lacking. Methods: We conducted a scoping review by adhering to the framework proposed by Arksey and O'Malley. We searched five electronic databases for literature published after 1990, using the key categories of "hospitals," "fires," and "education." As a result, we identified 15 relevant articles that met our inclusion criteria for the review. Results: Of the 15 articles, 12 had adopted a quasi-experimental design and the remaining 3 had employed a true experimental design. The majority of these studies (11 out of 15) were conducted in the United States, with 4 studies forming committees or teams dedicated to education. Simulation methods were used in 13 studies, while 2 studies had employed a combination of methods. All studies focused on first-response procedures based on RACE (Rescue, Alarm, Contain, Extinguish/Evacuation). Outcome measures included the learners' overall experience, performance in the educational settings, and performance in the field, with all studies reporting positive results following the educational interventions. Conclusion: Our review highlights the importance of multi-professional and multi-departmental educational strategies based on institutional-level initiatives for healthcare providers to create a safe hospital environment.

A Study on Forecasting the Repair Time Range of the Building Components in the Apartment Housing (공동주택 구성재의 예상수선시기 범위 설정 연구)

  • Lee Kang-Hee
    • Journal of the Korean housing association
    • /
    • v.17 no.2
    • /
    • pp.19-26
    • /
    • 2006
  • Building would be deteriorated with time elapse, influenced by its geographic situation, climate and other environmental conditions. In addition, the systematic maintenance could be provided to keep the resident a recent living condition. The existing breakdown maintenance will be changed into the preventive maintenance. The preventive maintenance is required to get the repair time, the repair scope and frequency. In this paper, it aimed at providing the repair time range over the building components, utilizing the relation between the determination curve and the performance recovery through repair. Results of this study are as follows : First, the forecast of the repair time over the building components could be calculated and equalized with the deterioration and performance degree. Second, the repair time range of building components would be provided into five categories and 3rd repair time. Results of this study will set up the long-term repair plan of building, and finally keep an housing condition comfortable.