• Annual Conference on Human and Language Technology
• /
• 1997.10a
• /
• pp.7-14
• /
• 1997

문서 범주화에서 문서의 내용에 따라 적합한 범주의 종류와 수를 찾는 문제를 해결하기 위해서는 문서 당 하나의 범주를 할당할 경우에 가장 좋은 성능을 보이는 모델이 효과적일 것이다. 그러므로, 본 논문에서는 문서 당 하나의 범주를 할당할 경우에 좋은 결과를 보이는 k-nearest neighbor 방법을 이용한다. 그리고 k-nearest neighbor 방법을 이용한 문서 범주화의 성능을 향상시키기 위해서, 문서 표현에 사용하는 단어들을 범주 자질의 성격을 갖는 단어들로 제한하는 방법을 제안한다. 제안한 방법은 Router 신문 일년치로 구성된 Router-21578 테스트 집합에서 breakeven point 82%라는 좋은 결과를 보였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.01a
• /
• pp.11-14
• /
• 2015

많은 블로그 제공 사이트는 블로그 포스트 작성자에게 미리 정의된 범주 (category)에 따라 포스트의 주제에 대하여 범주를 선택할 수 있는 환경을 제공한다. 그러나 블로거들은 작성한 포스트의 범주를 매번 수동으로 선택해야 하는 불편함이 있다. 이러한 불편함의 해결을 위해 블로그 포스트를 자동으로 분류해주는 기능을 제공한다면 블로그의 활용성이 증가할 것이다. 기존의 블로그 문서 분류의 연구는 각 범주의 고유 정보를 반영하는 것에 한계가 있었다. 이러한 문제를 해결하기 위해, 본 논문에서는 범주별 고유 정보를 반영한 어휘 가중치를 제안한다. 어휘 가중치의 분석을 위하여 범주별로 블로그 문서를 수집하고, 수집한 문서에서 어휘의 빈도와 문서의 빈도, 범주별 어휘빈도 등을 고려하여 새로운 지표인 CTF, CDF, IECDF를 개발하였다. 이러한 지표를 기반으로 기존의 Naive Bayes 알고리즘으로 학습하여, 블로그 포스트를 자동으로 분류하였다. 실험에서는 본 논문에서 제안한 가중치 방법인 TF-CTF-CDF-IECDF를 사용한 분류가 가장 높은 성능을 보였다.

• Proceedings of the Korean Society for Information Management Conference
• /
• 2014.08a
• /
• pp.95-98
• /
• 2014

본 연구는 지역명 범주에 속하는 용어들의 개념적 특성을 분석하고, 이를 토대로 다른 범주와의 관련도를 파악하여 지역명 범주 용어들을 중심으로 관계 속성들 사이의 논리적 연관성을 부여할 수 있는 모형 도출에 기반이 되는 기초 연구이다. 지역명 범주 용어 중 국가명에 한정하여 분석한 결과, 국가명 개념 속성 중심으로는 계층 구조 관계의 지역명 범주 용어들끼리 연관이 높으며, 전체 범주 용어들의 개념 속성 중심으로는 지역명 범주 용어가 지리적 위치로서의 의미로 주로 쓰이나, 행위의 주체 또는 객체의 의미나 시대의 개념으로도 많이 활용됨을 알 수 있었다. 국가명이 참조되는 개념 속성과 연관되어 활용되는 관계 속성의 경우의 일부는 참조하는 주요 개념 범주와 연관 관계를 토대로 논리적 의미 관계를 생각해볼 수 있는 것으로 나타났다.

• Journal of Korean Library and Information Science Society
• /
• v.50 no.1
• /
• pp.273-291
• /
• 2019

This study aims to extend fundamental categories from Clause 11, "facet analysis" in International Standards for thesaurus(ISO 25964-1) by analyzing fundamental categories of Clause 11 and concept and their scope in a thesaurus of Clause 5. For to do this, the fundamental categories were established by adjusting partially and adding mental entities explicitly referencing the highest concepts(YAMATO which is the upper ontology of Mizoguchi, and ISO 2788) and existing fundamental categories(PMEST, FRBR group 3 entities, 13 categories in CRG). Also, established fundamental categories were reorganized and structured based on concreteness/abstraction of PMEST in Ranganathan and independence/dependence of YAMATO in Mizoguchi. And the upper categories were divided into independent and dependent entities. Under these entities 28 criteria are included in the independent ones and 2 criteria in the dependent ones. In the further study, the result of this study can be expected to reuse and refer as controlled vocabulary in the field like classification, taxonomies and thesauri where expected to utilize fundamental categories and as the high-level concept when constructing an ontology for information retrieval.

• Korean Journal of Cognitive Science
• /
• v.30 no.1
• /
• pp.31-50
• /
• 2019

The present study is designed to investigate the effect of perceptive, emotional and semantic distinctiveness on retrieval-induced forgetting(RIF). Experiment 1 was designed to construct a category and category list for RIF experimental paradigm and to investigate the effects of perceptual distinctness on retrieval-induced forgetting. It was used for the list consisting of the six categories and six words in each category list. In controlled conditions, all the stimuli were presented in black and Gothic. In contrast, perceptual distinctiveness conditions, half of the category list were presented in red and Gungseoche. RIF was observed in all conditions. Experiment 2 was designed to investigate the effects of semantic and emotional distinctiveness on retrieval-induced forgetting. In neutral conditions, adjectives related to items were added. In the emotional distinctiveness condition, half of the items in the category were manipulated in such a way as to add the negative adjectives. In the semantic distinctiveness condition, half of the items in the category were manipulated in such a way as to add the inappropriate adjective. As a result, RIF occurred in the neutral condition, but RIF did not occur in both the emotional discrimination condition and the semantic discrimination. These results suggest the possibility that the RIF will not occur when the distinctiveness occurs within a categorical relationship.

• Journal of Korean Library and Information Science Society
• /
• v.35 no.2
• /
• pp.135-154
• /
• 2004

This study was investigated to the geographical categories and arrangement of library classifications and public organizations in ender to understand the geographical category comprehension and canon from the north-west to south-east in Korean library classification and then to develop a Korean geographical classification system Research method was the examination of different type of library classification (DDC, NDC, LCC) and geographical categorizations of the public organizations. The results show that there was not a principle of categorization and arrangement for geographical information. This study proposes to the three hierarchical structures and radial category arrangement from center-area to side area for developing Km geographical classification.

• Korean Journal of Cognitive Science
• /
• v.12 no.4
• /
• pp.29-39
• /
• 2001

Kaplan(2000) reported that instances were categorized more accurate in the aligned condition than in the non-aligned condition irrespective of similarity between instances[16]. This study investigated wether Kaplan(2000)\\`s results could be explained by stimulus types she used and alignment effects in categorization were due to selective attention to aligned attributes. In Experiment 1. I examined whether attribute alignment produced significant effects on similarity and categorization and aligned attributes were recalled more than non-aligned ones. Results showed that instances were rated more similar and categories were learned more rapidly in the aligned condition than in the non-aligned condition. It can be explained that categories are learned rapidly in the aligned condition because attribute alignment increases within-category similarity. But. the result that aligned attributes were recalled more than non-aliened ones in the attribute recall test implies that alignment effects in categorization can be independent of similarity between instances partially. In Experiment 2. I used equal numbed of attributes defining two categories and instructed subjects to pay their attention to categorization-relevant dimensions only. Results showed that dimension instruction facilitated category learning in the non-aligned condition only but categories were learned more rapidly in the aligned condition than in the non-aliened condition irrespective of instruction types. In conclusion. attribute alignment in categorization may facilitate paying selective attention to categorization-relevant attributes.

• 한국노년학
• /
• v.32 no.2
• /
• pp.611-629
• /
• 2012

This research examines the participation experience of a Focusing group program dealing with elderly depression, with the aim of providing a groundwork for developing and improving such programs. Qualitative analysis is performed based interviews conducted with 9 elderly clients with depression symptoms that participated in a 10 week Focusing group program. The results summarize into 3 themes, 3 categories per theme, and 3 sub-categories per category. The first theme is "program effectiveness" with three associated categories: Category 1) The reduction effect of the depression ; Category 2) Increased self-awareness, self-respect and acceptance experience; Category 3) Everyday life changes. The second theme is "expectation and difficulties of the program", with 3 related categories: Category 1) Expectation of program; Category 2) Difficulties of program; Category 3) Personal difficulties. The third theme is "program enhancement methods" with 3 affiliated categories: Category 1) Differentiation with other programs; Category 2) Improvement of program; Category 3) Post program direction. Based on the results of the analysis, the effectiveness of the Focusing group program for elderly depression is examined and suggestions are provided for further program improvements and applications.

• Asia Marketing Journal
• /
• v.13 no.4
• /
• pp.1-27
• /
• 2012

Extending the findings of Rajagopal and Burnkrant (2009), this research examines the moderating role of consumers' goals (i.e., head category-relevant goal vs. modifier category-relevant goal) in the effects of two different interpretation strategies (i.e., relational interpretation vs. property interpretation) on product beliefs and attitudes toward hybrid products. In the current research, we make two predictions. First, we predict that both head category and modifier category beliefs will be higher under property interpretations than under relational interpretations in the modifier category-relevant goal priming conditions, whereas there will be no significant differences between each product category beliefs across the two interpretation conditions in the head category-relevant goal priming conditions. Second, we predict that attitudes toward hybrid products will be higher under property interpretations than under relational interpretations in the modifier category-relevant goal priming conditions, whereas there will be no significant differences between the attitudes toward hybrid products across the two interpretation conditions in the head category-relevant goal priming conditions. These predictions are tested and confirmed in two experiments. Finally, we discuss theoretical and practical implications of our findings and develop directions for future research.

• Information Systems Review
• /
• v.24 no.4
• /
• pp.23-40
• /
• 2022

Although Support Vector Machine(SVM) has been used in various fields such as bankruptcy prediction model, the hyperplane learned by SVM in class imbalance problem can be severely skewed toward minority class and has a negative impact on performance because the area of majority class is expanded while the area of minority class is invaded. This study proposed optimized uneven margin SVM(OPT-UMSVM) combining threshold moving or post scaling method with UMSVM to cope with the limitation of the traditional even margin SVM(EMSVM) in class imbalance problem. OPT-UMSVM readjusted the skewed hyperplane to the majority class and had better generation ability than EMSVM improving the sensitivity of minority class and calculating the optimized performance. To validate OPT-UMSVM, 10-fold cross validations were performed on five sub-datasets with different imbalance ratio values. Empirical results showed two main findings. First, UMSVM had a weak effect on improving the performance of EMSVM in balanced datasets, but it greatly outperformed EMSVM in severely imbalanced datasets. Second, compared to EMSVM and conventional UMSVM, OPT-UMSVM had better performance in both balanced and imbalanced datasets and showed a significant difference performance especially in severely imbalanced datasets.