• Korean Journal of Plant Resources
• /
• v.35 no.4
• /
• pp.471-501
• /
• 2022

This study was carried out on 455 forest wetlands of south Korea for which an inventory was established through value evaluation and grade. Correlation analysis was conducted to find out the correlation between the types and grades of forest wetlands and 23 evaluation factors in four categories: vegetation and landscape, material circulation and hydraulics·hydrology, humanities and social landscape, and disturbance level. Through the improvement of types and grades of forest wetlands, it is possible to secure basic data that can be used in setting up conservation measures by preparing standards necessary for future forest wetland conservation and restoration, and to found a systematic monitoring system. First, between the type of forest wetland and size and accessibility showed a positive correlation, but the remaining items were analyzed to have negative or no correlation. In particular, it was found that there was no negative correlation or no correlation with the grades of forest wetland. Moreover, it was found that there was a very strong negative correlation with the weighted four category items. Thus, it is judged that improvement is necessary because there is an error in the weight or adjust the evaluation criteria of the value evaluation item, add an item that can increase objectivity. Especially, in the case of forest wetlands, the ecosystem service function due to biodiversity is the largest, so evaluation items should be improved in consideration of this. Therefore, it can be divided into five categories: uniqueness and rarity (15%), wildlife habitat (15%), vegetation and landscape (35%), material cycle·hydraulic hydrology (30%), and humanities and social landscape (5%). It will be possible to propose weights that can increase effectiveness.

• The KIPS Transactions:PartB
• /
• v.16B no.6
• /
• pp.497-502
• /
• 2009

In this paper, we introduce an automatic product feature extracting system that improves the efficiency of product review analysis. Our system consists of 2 parts: a review collection and correction part and a product feature extraction part. The former part collects reviews from internet shopping malls and revises spoken style or ungrammatical sentences. In the latter part, product features that mean items that can be used as evaluation criteria like 'size' and 'style' for a skirt are automatically extracted by utilizing term statistics in reviews and web documents on the Internet. We choose nouns in reviews as candidates for product features, and calculate degree of association between candidate nouns and products by combining inner association degree and outer association degree. Inner association degree is calculated from noun frequency in reviews and outer association degree is calculated from co-occurrence frequency of a candidate noun and a product name in web documents. In evaluation results, our extraction method showed an average recall of 90%, which is better than the results of previous approaches.

• Design Convergence Study
• /
• v.18 no.6
• /
• pp.27-39
• /
• 2019

The purpose of this study is to develop design guidelines for developing innovative products utilizing the characteristics of creative consumers(cresumers) who create alternative usages of consumer products, and to suggest effective way of applying the guidelines in design process. To do this, the study conducted the analysis on the characteristics and types of cresumers and in-depth interviews and observations of actual cresumers. The study was able to initially come up with 131 items in 7 categories through keywordization and affinity analysis of prior outcomes, and reorganized and modified those to 52 items in 4 categories. After verification of co-relationship and credibility analysis with 2 evaluation sessions (7-scale SD), the study has reached 4 categories - product concept development, product development, product use environment, and market possibility - with 46 guideline items. The study developed a design toolkit to effectively apply the guidelines in the product development process. The toolkit consists of three chapters and provides explanations and related examples to clarify the categories and individual guidelines, designed in the form of cards and booklet.

• Journal of KIISE:Databases
• /
• v.32 no.3
• /
• pp.297-303
• /
• 2005

When confronted with a query, question answering systems endeavor to extract the most exact answers possible by determining the answer type that fits with the key terms used in the query. However, the efficacy of such systems is limited by the fact that the terms used in a query may be in a syntactic form different to that of the same words in a document. In this paper, we present an efficient semantic query expansion methodology based on a question category concept list comprised of terms that are semantically close to terms used in a query. The semantically close terms of a term in a query may be hypernyms, synonyms, or terms in a different syntactic category. The proposed system constructs a concept list for each question type and then builds the concept list for each question category using a learning algorithm. In the question answering experiments on 42,654 Wall Street Journal documents of the TREC collection, the traditional system showed in 0.223 in MRR and the proposed system showed 0.50 superior to the traditional question answering system. The results of the present experiments suggest the promise of the proposed method.

• Journal of Distribution Research
• /
• v.10 no.4
• /
• pp.1-22
• /
• 2005

The major objectives of this study are twofold. The first is to discover which in-store tactics influence store performance when a retailer implements category management in variety enhancer and fill-ins categories. The second is to analyze how and why specific in-store tactics achieve better or worse performance than other in-store tactics across categories. The data were collected using scanner data and direct observations in 'A' discount store which is one of the representative discount stores in Korea. The in-store tactics were measured by product assortment, temporary price discount, price and non-price promotion, and shelving. The store performance was measured by sales and gross margin return on inventory investmant(GMROI). Empirical results analyzed by multiple regression were as follows: In variety enhancer category, the significant factors affecting sales were product assortment, temporary price discount, price promotion, and shelving. Non-price promotion also influenced GMROI positively but product assortment impacted on GMROI negatively. In fill-ins category, the significant factors affecting sales and GMROI were product assortment and shelving. However, the other factors such as temporary price discount, price promotion, and non-price promotion had no significant influence on both sales and GMROI. This paper presents a number of theoretical and managerial implications of the empirical results and concludes by addressing limitations and future research directions.

• Journal of Korea Society of Industrial Information Systems
• /
• v.20 no.3
• /
• pp.37-49
• /
• 2015

To develop the contents making profit, we propose the program operation methods that can apply the needs of students and the demand of the times, and consider the capacity of the operating agency overall. First, we suggest the distance training motivation, the effective content type, appropriate interaction ratio, the effective teaching and learning methods and the assessment methods. Second, we suggest the development strategy of educational contents, assess quantitative the demand of students and the will of teacher overall, measure the potential success. Third by applying the strategies in the educational field, we product the 12 major development field. These fields are divided into categories A and B, category A is the very high success field and category B is the high potential success field. By applying the proposed strategy, You will select the most suitable contents here and now.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.3
• /
• pp.227-234
• /
• 2017

In this paper, we propose a system whereby one can automatically classifies categories based on image data of the products for a shopping mall platform. Many products sold within internet shopping malls are classified their category defined by the same use of product names and products. However, it is difficult to search by category classification when the classification of the product is uncertain and the product classified by the shopping mall seller judgment is different from the purchasing user judgment. We proposes classification and retrieval method by Deep Learning technique solely using product image. The system can categorize products by using their images and its speed and accuracy are quantified using test data. The performance is evaluated with the test data. In addition, its usability is tested with the participants.

• Proceedings of the Korean Information Science Society Conference
• /
• 2005.07b
• /
• pp.484-486
• /
• 2005

본 논문은 백과사전 도메인의 지식베이스 설계 및 통계기반 정보추출 방법을 이용한 속성정보 인식에 대하여 기술한다. 층 13개 카테고리로 구성된 백과사전에 대해 99개의 템플릿과 285개의 속성을 정의하였으며, 각 표제어의 추출 대상인 속성정보는 표제어를 설명하는 본문에서 통계기반 기계학습모델인 CRF(Conditional Random Fields)를 적용하여 추출하였다. 백과사전 카테고리 별로 균일하게 선정된 4천 5백 문서를 학습에 사용하였고 테스트 문서셋 500문서에 대해 속성인식률을 측정하였다. 성능 평가한 결과, $F1\;55.76\%\;(P\;74.89\%,\;R\;44.42\%)$의 성능을 나타내었다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.231-232
• /
• 2021

본 논문에서는 딥러닝 모델 중 Keras 기반 LSTM 모델과 KoGPT-2 모델을 이용하여 학습한 결과를 바탕으로 카테고리 별 키워드 기반의 배달 음식점 리뷰를 생성하는 방법을 제안한다. 데이터는 주로 맛, 양, 배달, 가격으로 구성되어 있으며 이를 카테고리 별로 구분하였다. 또한 새롭게 생성된 텍스트는 의미와 문맥을 판단하여 기존 리뷰 데이터와 비슷하게 구현하였다. 모델마다 성능을 비교하기 위해 정량적, 정성적 평가를 진행하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.914-917
• /
• 2021

IoT 기기는 취약한 아이디와 비밀번호 사용, 저사양 하드웨어 등 보안 취약점으로 인해 사이버 공격 진입점으로 이용되고 있다. 본 논문은 IoT 악성코드를 탐지하기 위한 op 코드 카테고리 기반 특징 표현을 제안한다. Op 코드의 기능별 분류 정보를 이용해서 n-gram 특징과 엔트로피 히스토그램 특징을 추출하고 IoT 악성코드 탐지를 위한 기계학습 모델 평가를 수행한다. IoT 악성코드는 기능 개선과 추가를 통해 진화하였으나 기계학습 모델은 훈련 데이터에 포함되지 않은 진화된 IoT 악성 코드에 대한 예측 성능이 우수하였다. 또한 특징 시각화를 이용해서 악성코드의 비교 탐지가 가능하다.