• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.30 no.4
• /
• pp.669-677
• /
• 2020

As the number of Internet users exploded, attacks on the web increased. In addition, the attack patterns have been diversified to bypass existing defense techniques. Traditional web firewalls are difficult to detect attacks of unknown patterns.Therefore, the method of detecting abnormal behavior by artificial intelligence has been studied as an alternative. Specifically, attempts have been made to apply natural language processing techniques because the type of script or query being exploited consists of text. However, because there are many unknown words in scripts and queries, natural language processing requires a different approach. In this paper, we propose a new classification model which uses byte pair encoding (BPE) technology to learn the embedding vector, that is often used for web attack payloads, and uses an attention mechanism-based Bi-GRU neural network to extract a set of tokens that learn their order and importance. For major web attacks such as SQL injection, cross-site scripting, and command injection attacks, the accuracy of the proposed classification method is about 0.9990 and its accuracy outperforms the model suggested in the previous study.

• Proceedings of the Korea Concrete Institute Conference
• /
• 1994.10a
• /
• pp.273-278
• /
• 1994

The shear behavior of simply supported reinforced concrete deep beams subject to concentrated loads has been scrutinized experimentally to verify the influence of the structural parameters such as shear span ratio, and the horizontal and vertical web reinforcements. A total of 27 specimens has been tested at the laboratory. In the tests all specimens have failed in shear causing inclined cracks from the load application points to the supports. The load bearing capacities have changed significantly depending on the shear span ratio. The effects of the vertical and horizontal reinforcements on the shear strength and crack initiation and propagation have been carefully checked and analyzed.

• Magazine of the Korea Concrete Institute
• /
• v.5 no.2
• /
• pp.151-161
• /
• 1993

본 논문은 전단철근을 갖지 않는 비교적 짧은 지간의 철근콘크리트 보에서 전단특성을 규명하고 균열전단강도와 극한전단강도를 예측하기 위한 것으로 총30개의 보를 4 series로 나누어 실험을 수행하였다. 실험의 변수는 콘크리트의 강도, 전단지간-유효높이의 비, 인장철근량등이며, 실험과정을 통해 파괴형상, 처짐, 전단강도등을 측정하였다. 실험결과로부터 콘크리트의 강도가 커지고 철근량이 많아질수록, 그리고 전단지간이 짧아질수록 철근콘크리트 보의 균열 및 극한전단강도가 증가됨을 밝혔다. 또한, 실험성과를 회귀분석하여 균열전단강도와 극한전단강도 추정식을 제안하였다. 제안된 추정식에 의한 계산값과 실험성과를 비교 검토하여 그 상관성을 확인하였다.

• Journal of Information Technology Services
• /
• v.17 no.2
• /
• pp.165-177
• /
• 2018

Recently, online commerces are becoming more common due to factors such as mobile technology development and smart device dissemination, and online review has a big influence on potential buyer's purchase decision. This study presents a set of analytical methodologies for understanding the meaning of customer reviews of products in online transaction. Using techniques currently developed in deep learning are implemented Hierarchical Attention Network for analyze meaning in online reviews. By using these techniques, we could solve time consuming pre-data analysis time problem and multiple topic problems. To this end, this study analyzes customer reviews of laptops sold in domestic online shopping malls. Our result successfully demonstrates over 90% classification accuracy. Therefore, this study classified the unstructured text data in the semantic analysis and confirmed the practical application possibility of the review analysis process.

• The Journal of Information Systems
• /
• v.29 no.3
• /
• pp.103-117
• /
• 2020

Purpose Recently, various issues related to toxic comments on web portal sites and SNS are becoming a major social problem. Toxic comments can threaten Internet users in the type of defamation, personal attacks, and invasion of privacy. Over past few years, academia and industry have been conducting research in various ways to solve this problem. The purpose of this study is to develop the deep learning modeling for toxic comments classification. Design/methodology/approach This study analyzed 7,878 internet news comments through CNN classification modeling based on Highway Network and OOV process. Findings The bias and hate expressions of toxic comments were classified into three classes, and achieved 67.49% of the weighted f1 score. In terms of weighted f1 score performance level, this was superior to approximate 50~60% of the previous studies.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.04a
• /
• pp.791-794
• /
• 2015

최근 정보화 시대에 들어 개인 사진을 SNS 및 클라우드 서비스에 업로드 한다. 하지만 각각의 사진 데이터만 클라우드 및 SNS에 업로드 되며 사진 검색에 있어 불편한 부분이 많다. 따라서 사진에 태그 분류 서비스 및 카테고리를 자동으로 부여해 업로드를 완료한 후 자동 사진 정리 및 사진 검색의 편리함을 도모하고자 한다.

• International Journal of Advanced Culture Technology
• /
• v.7 no.2
• /
• pp.227-237
• /
• 2019

In this paper, we have developed a Retina Healthcare Service System through which the patient himself/herself can manage his/her retina health. In the case of conventional portable ophthalmic cameras, patients cannot check their eye health on their own because most of them are used by doctor in environments where ophthalmography cannot be performed properly. This system consists of web, app and camera modules, and when a patient mounts a camera module for fundus photography on his / her smart phone and then photographs his / her fundus through the app, the image is transmitted to a server, and the transmitted image reads the fundus the patient's fundus image status in the fundus image reading model learned using deep learning. When the doctor expresses his/her opinions about the patient 's eye condition based on the reading result and the fundus photograph, the patient can check through the app and judge whether to receive ophthalmologic treatment.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.07a
• /
• pp.187-188
• /
• 2021

본 논문에서는 실시간으로 web-cam을 이용해, 후숙과일의 불량 여부를 판단, 분류하고 불량이 없는 후숙과일의 이미지 분석을 통하여 숙성도 예측하는 시스템을 소개한다. 실시간 다중 객체인식에 탁월한 yolo모델을 활용해, 과일의 불량여부 판단 후 분류하고, 이미지를 획득한 뒤, k-mean clustering 알고리즘을 이용해, 이미지를 segmentation 한다. segmentation된 이미지에 grabcut 알고리즘의 foreground-extraction을 사용해 배경 제거를 한 뒤, cluster의 중심색상값 색상값의 면적%, 전체 면적을 이용해 현재 숙성도를 계산하고 이를 이용해 과일의 후숙 시간 데이터와 비교, 숙성이 완료될 시간을 예측한다. 기존 수작업으로 이루어지고 있는 과일의 분류작업의 인력 감소 및 정확성을 높일 수 있는 알고리즘을 제안한다.

• Journal of The Korean Association of Information Education
• /
• v.11 no.2
• /
• pp.147-155
• /
• 2007

Paper & pencil test provides limited informations about student. So many web-based computerized test systems are developed and tried. But These systems have drawbacks that cannot visit previous or next problems freely. So These systems cannot get deep informations such as elapsed time of each problem, history of answer modification, the number of visit time of each problem. These deep informations are very important in educational evaluation area. Therefore we develop educational evaluation system can get these deep informations including basic informations. And then we apply paper & pencil test and our system to 67 students, and execute survey. The reason of survey is to ascertain availabilty of our system.

• The Journal of the Korea Contents Association
• /
• v.6 no.3
• /
• pp.12-19
• /
• 2006

In this paper, we describe a deep analysis of question for question answering system. It is difficult to offer the correct answer because general question answering systems do not analyze the semantic of user's natural language question. We analyze user's question semantically and extract semantic features using the semantic feature extraction grammar and characteristics of natural language question. They are represented as semantic features and grammatical morphemes that consider semantic and syntactic structure of user's questions. We evaluated our approach using 100 questions whose answer type is a person in the web. We showed that a deep analysis of questions which are comparatively short but enough to mean can analysis the user's intention and extract semantic features.