• 영어어문교육
• /
• 제3호
• /
• pp.55-70
• /
• 1997

This paper aims at showing firstly, that the parameterized Universal Grammar is also accessible in second language acquisition based on the data found in the English Agreement acquisition by Korean speakers and secondly, that the theoretical claim that the acquisition of Agreement is related to the Subject Drop phenomenon can be supported by the second language acquisition data. The functional category AGRP which is crucial here can be parameterized according to its features with Korean of - Agr and English of + Agr. Two groups of 40 each were tested and the result shows that the parameter resetting is possible in terms of the Agreement feature. In addition to Agreement test, three more tests for the distribution of Subject Drop, Subject Raising and NPI (Negative Polarity Items) were conducted in order to find the correlation among those grammatical phenomena. The result is that the acquisition of Agreement and the Subject Drop possibility are correlated but that they are not related to the obligatory Subject Raising process. Finally, NPI distribution test which is supposedly related to the Subject Raising turned out to carry little information since the average grades from both groups were very low.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.399-410
• /
• 2021

Aiming at the problem that complex external variables in natural scenes have a greater impact on facial expression recognition results, a facial expression recognition method based on two-stream convolutional neural network is proposed. The model introduces exponentially enhanced shared input weights before each level of convolution input, and uses soft attention mechanism modules on the space-time features of the combination of static and dynamic streams. This enables the network to autonomously find areas that are more relevant to the expression category and pay more attention to these areas. Through these means, the information of irrelevant interference areas is suppressed. In order to solve the problem of poor local robustness caused by lighting and expression changes, this paper also performs lighting preprocessing with the lighting preprocessing chain algorithm to eliminate most of the lighting effects. Experimental results on AFEW6.0 and Multi-PIE datasets show that the recognition rates of this method are 95.05% and 61.40%, respectively, which are better than other comparison methods.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2019년도 제59차 동계학술대회논문집 27권1호
• /
• pp.123-124
• /
• 2019

This paper proposes an improved deep learning method based on small data sets for animal image classification. Firstly, we use a CNN to build a training model for small data sets, and use data augmentation to expand the data samples of the training set. Secondly, using the pre-trained network on large-scale datasets, such as VGG16, the bottleneck features in the small dataset are extracted and to be stored in two NumPy files as new training datasets and test datasets. Finally, training a fully connected network with the new datasets. In this paper, we use Kaggle famous Dogs vs Cats dataset as the experimental dataset, which is a two-category classification dataset.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1991년도 한국자동제어학술회의논문집(국제학술편); KOEX, Seoul; 22-24 Oct. 1991
• /
• pp.1570-1575
• /
• 1991

The feasibility of using an adaptive resonance network (ART2) with unsupervised learning capability for too] wear detection in turning operations is investigated. Specifically, acoustic emission (AE) and cutting force signals were measured during machining, the multichannel AR coefficients of the two signals were calculated and then presented to the network to make a decision on tool wear. If the presented features are significantly different from previously learned patterns associated with a fresh tool, the network will recognize the difference and form a new category m worn tool. The experimental results show that tool wear can be effectively detected with or without minimum prior training using the self-organization property of the ART2 network.

• 한국멀티미디어학회논문지
• /
• 제20권7호
• /
• pp.1065-1072
• /
• 2017

As smart devices and social network services (SNSs) become increasingly pervasive, individuals produce large amounts of data in real time. Accordingly, studies on unstructured data analysis are actively being conducted to solve the resultant problem of information overload and to facilitate effective data processing. Many such studies are conducted for filtering inappropriate information. In this paper, a feature-weighting method considering SNS-message features is proposed for the classification of short text messages generated on SNSs, using Markov logic networks for category inference. The performance of the proposed method is verified through a comparison with an existing frequency-based classification methods.

• 산업공학
• /
• 제13권1호
• /
• pp.29-36
• /
• 2000

Many organizations have applied the traditional Information Strategy Planning methodologies to planning knowledge management. This approach, however, cannot sufficiently realize the strategic value of various kinds of knowledge. To overcome above problems, this research classifies knowledge according to its features and identifies characteristics of each knowledge category. And then, structured Knowledge Strategy Planning methodology, $P^2-KSP$, is proposed. With $P^2-KSP$, an organization can identify organizational knowledge and develop KM architecture from both perspectives of the performance measurement system and business processes.

• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 1997년도 제9회 한글 및 한국어 정보처리 학술대회
• /
• 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.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2001년도 봄 학술발표논문집 Vol.28 No.1 (B)
• /
• pp.364-366
• /
• 2001

문서 범주화는 전자 정보환경에서 매우 유용한 정보처리 도구로서, 다양한 문서 범주화 기법 및 성능향상을 위한 연구들이 지속적으로 이루어지고 있다. 그러나, 대부분의 연구들은 문서 범주화의 대상이 되는 단어 자질 공간의 차원축소 문제에만 집중되었을 뿐, 학습단계에 큰 영향을 미치는 다범주 단어 자질의 범주 모호성은 고려하지 않았다. 본 연구에서는, 다범주 자질의 범주 모호성을 해소함으로써 문서 범주화의 성능향상을 유도하는 범주 모호성 해소 가중치 W를 제시하고 이를 실험을 통해 증명하였다. 실험에서는 Naive Bayes 분류기와 가중치 W를 적용한 Naive Bayes-W 분류기를 직접 구축하여 문서 범주화의 성능향상 여부를 비교하는데 사용하였다. 도출된 실험결과를 통해, 가중치 W는 현재의 분류기가 가지고 있는 자질 표현의 범주 모호성이라는 단점을 보완하고 분류기의 성능향상을 유도함으로써 정보검색시스템의 검색효율을 높이는 데 활용될 수 있음일 증명되었다.

• 한국가정과교육학회지
• /
• 제10권2호
• /
• pp.87-99
• /
• 1998

The purposes of the study are to see as to how much of changes has occurred in the newly revised high school Home Economics Textbooks required by the 6th Curriculum and to seek future policy directions which can be improved even better, if any, for further revision. To implement this proposal, the study employs 6 different textbooks published in 1996 geared for the 6th Curriculum for data base. The components of the contents divides into two major parts. The first part faUs into a category of holistic analysis of the textbooks, which gives a general perspective of the changes in the textbook revised. It includes those areas of changes in the total number of pacover page, illustrations, table of contents, appendices, unit cover, introductory remarks, summary and evaluation, etc. It also covers the information on authors of textbook. The latter part consists of several specific areas of major changes occurred in the textbooks for the 6th Curriculum. It covers such areas of changes in the composition of category and its weigh laboratory experiences, and illustrations and tables. To conclude the study, there found followings key features : 1) The new textbooks have adopted more of varieties in physical appearances and contents per se as to the older ones. And yet, there still found a weakness in numbers of authors participated to a full reflection of contents variety. In addition, the background of authors reflected on an unrealistic idealism which lacks the on-site information provided by the school teachers. 2) The content area consists of 3 major areas of Human Development & Family Relationship. Food and Nutrition. Clothing and Textiles with ratio of 23.8%, 23.1%, and 21.9% respectively. 3) Human Development & Family Relationship area ranked first with 71 times in the numbers of laboratory activities for each content area. followed by Food and Nutrition with 56 and Family Resorces & Consumerism with 54. These increased numbers of laboratory activities in textbooks represent the changing contents of Home Economics and imply the desirable and positive directions of Home Economics education in Korea. numbers of laboratory activities for each content area. followed by Food and Nutrition with 56 and Family Resorces & Consumerism with 54. These increased numbers of laboratory activities in textbooks represent the changing contents of Home Economics and imply the desirable and positive directions of Home Economics education in Korea.

• 한국실내디자인학회논문집
• /
• 제16권2호
• /
• pp.79-86
• /
• 2007

The recent works of architecture and interior design take on too various aspects to be classified as a trend or tendency. Many subjects, seeking the superiority between form and function in space which was the issue of Modernism, got to face the works which transcend the limitation of architecture due to the development of a variety of media and technology these days. In result, we could see the works which are to be Included into the category of art rather than architecture. This study starts from the thought that the common aspects of these works are based on the strangeness to pursue uniqueness, peculiarities or novelty. 'The subject of this study, $'D\acute{e}paysement'$, creates a new relation, disconnected from customs, to an architecture which became familiar through defragment, fixation and abstraction by the habitual relation and makes it reborn as an architecture which gives us a fantasy and a pleasure, by converting into the shape and image which are different from its original shape when a stimulus is given to an architecture. This study starts from the artistic aspect which has a close correlation with the architecture. The subject, $'D\acute{e}paysement'$ could be historically considered in two aspects - $D\acute{e}paysement$ of Surrealism which developed in paintings and Defamiliarization of Russian Formalism in the area of literature. This study is carried in a category of $D\acute{e}paysement$, focusing on the visual aspect of the architecture. This study is composed as follows: In chapter 2, a theoretical research is performed according to the definition and meaning of the $D\acute{e}paysement$ expression techniques, its origination and evolution, and type classification by the features of expression techniques in the era of Surrealism when $D\acute{e}paysement$ developed most vigorously in the area of art. In chapter 3, the system which formalizes the expression method of the techniques is presented, focusing on the case analysis of the artistic works according to the types of $D\acute{e}paysement$. In chapter 4, the application of the expression techniques of $D\acute{e}paysement$ is discussed by conducting cases analysis of the modern interior design.