Earticle

이 논문의 목적은 국립극장 극장레퍼토리 구축 사례를 국립무용단 세 작품, <춤, 춘향>, <Soul, 해바 라기>, <항연(饗宴)>을 통해 분석함에 있다. 연구방법인 ‘극장레퍼토리 구축 순환구조’를 통해 극장레퍼 토리 구축 사례를 진단한다. 논의 결과를 종합하면 다음과 같다. <춤, 춘향>은 국가브랜드공연으로서 국립극장의 고유성, 상징성, 정체성 확립에 기여한다. 작품 시, 캐스팅 등 인적 변화 시도, 다양한 유형 의 공연 시행을 통해 환류 기능을 충실히 수행했다. 국립무용단 스테디공연인 <Soul, 해바라기>는 국립 무용단 ‘다시 보고 싶은 공연 1위’라는 관객 평가가 정책적 측면에서 작동할 수 있는 동인이 된다. 높은 예술성과 객석점유율 등 정량적 지표 결과는 경영성을 충족하는 요소로 작용한다. <향연>은 전통춤 현대적 수용의 전형성을 보여준다. 원형 기반 재창조와 재해석 작업의 유의미한 성과는 극장레퍼토리 구축 토대가 될 수 있다. 세 작품 사례 분석을 통해 볼 때, 극장레퍼토리 3요소인 대표성, 예술성, 지속성은 ‘정책 결정’, ‘제작 환경’, ‘환류’로 수렴되고 확산됨을 확인했다.

The purpose of this dissertation is to analyze the National Theatre of Korea’s theater repertoire foundation using Dance, Chun-Hyang; Soul, Sunflower; and Hyang-Yeon. The analysis will be performed using the ‘the circulation structures of establishing theater repertories’ method. The results of the analysis are as follows. As a national brand performance, Dance, Chun-Hyang is a National Theatre of Korea’s unique, symbolic, identity establishment contribution. Through the production and casting, as well as extraordinary attempts in casting, various performance techniques are used to successfully achieve feedback system. The National Dance Company of Korea’s steady seller Soul, Sunflower was selected as a National Theatre of Korea’s ‘most wanted to be seen again’ performance by the audience, which can influence the company’s management policy. With its high artistic quality and box-office popularity, the piece shows an exemplary case that satisfy administrative requirements. Hyang-Yeon represents a traditional dances that been modernized. The reinvention and reinterpretation of traditional dance in this piece produce meaningful outcome that would provide the structural foundation of the repertoire. When analyzing works of art, the theater repertoire’s 3 aspects of representativeness, artistic value, and continuity has been seen to aid in ‘policy decisions’, ‘production environment’, and ‘feedback system’ for convergence and divergence.

최근 우리 사회의 많은 관심 대상인 청소년 전자게임 문화 현황에 대하여 고찰하고, 몇몇 청소년들이 게임 중독, 아이템불법매매 등으로 여러 반사회적 문제를 야기 시켰다. 특히 우리 사회에서 거의 모든 청소년 들은 게임을 즐기고 있다. (97.8%.게임백서 2008) 특히 한국사회에서 게임에 대한 본질적인 이해 부족으로 ‘청소년 게임문화 ’에 대하여 학교나 가정에서 ‘억제’ 중심으로 접근하여 많은 갈등을 야기 하였다. 그러나 바로 온라인게임이라는 매체의 특수성과 다양한 연령대를 아우르고 있는 특징 때문에 때로는 게임이 제공하고 있는 콘텐츠와 상관없이 여러 사용자들의 의해 우리의 청소년들이 유해한 환경들에 노출되게 되는 것이다. 쉬운 예로 게임 내 욕설을 들 수 있는데 이는 온라인 게임에서 공통적으로 대두되고 있고, 또 잘 알려진 문제점들이기도 하다. 따라서 이 논문에서는 온라인 게임에서 청소년들에게 윤리적으로 부정적인 영향을 미칠 수 있는 요소를 고찰하고 동시에 게임 내 사회적 환경에 초점을 맞추어 게임 내 비윤리적인 게임 플레이가 청소년들에게 미칠 수 있는 부정적인 영향들을 조명하고 온라인게임 개발사가 이런 게임 내 윤리적 사회 환경 조성에 있어 책임이 있음을 시사하며 더불어 해결 대책 마련의 필요성을 제안한다.

There are a lot of culturally affecting society as well as game industry by a fast development of an electronic game. From the latter half of 1990's, existing game method changed from independent (stand-alone) game methods to on-line games types. And It made a interchange of human various form and cultural formation in games. However it is not difficult to find these elements in online games because of the media of expression and variety in age. The main concern of this is that youth can be exposed to unhealthy environment in games without reference to game ratings. In this paper we will accentuate the significance of designing ethical social environment in games and suggest remedies for the issues regarding online computer games indicating that the developers and society have responsibility to make an effort in designing ethical social environment in games for youth. To concrete this suggestion, we will also define unfair behavior that could harm youth in on line-games focusing on in-game social environment.

전체 이용가 게임이란 말 그대로 모든 연령대가 이용 가능한 게임이란 뜻이다. 게임물 등급위원회에서는 선정성, 폭력성, 반사회적 묘사, 부적절한 언어, 사행성의 다섯 가지 요소를 등급분류의 기준으로 삼고 있다. 그러나 전체 이용가 온라인 게임에서 이 다섯 가지 요소에 해당되는 문제점들을 찾아보기란 매우 쉬운 일이다. 바로 온라인이라는 매체의 특수성과 다양한 연령대를 아우르고 있는 전체 이용가라는 게임등급 때문에 때로는 게임이 제공하고 있는 콘텐츠와 상관없이 여러 사용자들의 의해 아동들이 유해한 환경들에 노출되게 되는 것이다. 쉬운 예로 게임 내 욕설을 들 수 있는데 이는 게임의 등급과 상관없이 온라인 게임에서 공통적으로 대두되고 있고 또 잘 알려진 문제점들이기도 하다. 따라서 이 논문에서는 전체 이용가라는 게임등급에 한하여 아동들에게 윤리적으로 부정적인 영향을 미칠 수 있는 요소들로 범위를 좁히는 동시에 게임 내 사회적 환경에 초점을 맞추어 게임 내 비윤리적인 게임 플레이가 아동들에게 미칠 수 있는 부정적인 영향들을 조명하고 전체 이용가 게임 개발사가 이런 게임 내 윤리적 사회 환경 조성에 있어 책임이 있음을 시사하며 더불어 해결 대책 마련의 필요성을 알리려고 한다.

Game rating board considers five factors; sexuality, violence, antisocial description, inappropriate language, gambling in assigning a rating. However it is not difficult to find these elements in E rated online games because of the media of expression and variety in age. The main concern of this is that juvenile can be exposed to unhealthy environment in games without reference to game ratings. In this paper we will accentuate the significance of designing ethical social environment in games and suggest remedies for the issues regarding online computer games indicating that the developers and society have responsibility to make an effort in designing ethical social environment in games for juvenile. To concrete this suggestion, we will also define unfair behavior that could harm children and youth in E rated online computer games focusing on in-game social environment.

Background: In modern society, the use of computers accounts for a large proportion of our daily lives. Although substantial research is being actively conducted on musculoskeletal diseases resulting from computer use, there has been a recent surge in interest in improving the working environment for prevention. Objects: This study aimed to examine the effects of posture correction feedback (PCF) on changes in neck posture and muscle activation during computer typing. Methods: The participants performed a computer typing task in two sessions, each lasting 16 minutes. The participant's dominant side was photographed and analyzed using ImageJ software to verify neck posture. Surface electromyography (EMG) was used to confirm the participant's cervical erector spinae (CES) and upper trapezius muscle activities. The EMG signal was analyzed using the percentage of reference voluntary contraction and amplitude probability distribution function (APDF). In the second session, visual and auditory feedback for posture correction was provided if the neck was flexed by more than 15° in the initial position during computer typing. A 20-minute rest period was provided between the two sessions. Results: The neck angle (p = 0.014), CES muscle activity (p = 0.008), and APDF (p = 0.015) showed significant differences depending on the presence of the PCF. Furthermore, significant differences were observed regarding the CES muscle activity (p = 0.001) and APDF (p = 0.002) over time. Conclusion: Our study showed that the feedback system can correct poor posture and reduces unnecessary muscle activation during computer work. The improved neck posture and reduced CES muscle activity observed in this study suggest that neck pain can be prevented. Based on these results, we suggest that the PCF system can be used to prevent neck pain.

Objective: This study aimed to assess the impact of using the LED light electromyographyfeedback system (EMG-light) during Pilates exercises in women to maintain consistent muscle contraction in the abdominal external oblique muscles and reduce muscle contraction in the upper trapezius muscles. This study compared Pilates training using the EMG-light with Pilates training only in healthy women for 4 weeks. Design: This study was conducted as a cross-sectional study. Methods: A total of 17 healthy women were divided into an experimental group (n=9) and a control group (n=8). Both groups performed Pilates exercises as assigned, twice per week for four weeks, with each session lasting fifty minutes. The experimental group were used the EMG-light feedback system during pilates exercise while the control group did not use EMG-light. We used cervical vertebral angle (CVA), shoulder tilt angle to evaluate neck posture and standing balance with closed eye for 30s before and after exercise. Results: The control group exhibited a significant change in CVA (p<0.05). Both groups showed significant changes in shoulder tilt, center of pressure (COP) path-length, and COP velocity during eyes-closed conditions (p<0.05). However, there were significant differences between the experimental and control groups in terms of CVA, shoulder tilt, COP path-length, and COP velocity. Conclusions: This study demonstrated that Pilates exercises had positive effects on shoulder posture and balance. The use of EMG-light provided real-time visual feedback on muscle contraction during Pilates exercise. However, the experimental group did not show significant improvements compared to the control group, which performed Pilates exercises without feedback.

The linearization of an input/output controller has been designed for an input time delay nonlinear time discretized nonlinear system. The time discretized nonlinear model has been obtained based on Taylor-Lie series expansion method and zero order hold assumption. The resulting control algorithm enables the time delay nonlinear system control, while the continuous time controller cannot handle a time delay nonlinear system due to its infinite dimensionality. The performance of the proposed controller is evaluated by using two different case studies: a Van Der Pol equation and a Continuous Stirred Tank Reactor (CSTR) system that all exhibit nonlinear behavior and input time delay. For all the case studies, the results validate the proposed methods.

Purpose: To find the effect of visual feedback of head angle with using a mobile posture-aware system on craniocervical angle and neck and shoulder muscles fatigue for preventing or decreasing the forward head posture. Methods: Twenty-four healthy young adults in Chungbuk to participate in this study. The subjects started to watch a movie clip for 10 minutes with visual feedback in $0^{\circ}$, $30^{\circ}$, and $60^{\circ}$ of head angles. During the task, surface electromyography (EMG) was used to collect data from the upper trapezius (UT), sternocleidomasetoid muscle (SCM), cervical erecter spinae (CES) during watching the smartphone. Craniocervical angles were measured using a sagittal-view photograph of the subject in a sitting posture. A one-way repeated analysis of variance with a significant level of 0.05 used for statistical analysis. Results: Craniocervical angle with $0^{\circ}$ visual feedback was significantly greater than $30^{\circ}$ and $60^{\circ}$. Craniocervical angle with $30^{\circ}$ visual feedback was significantly greater than $60^{\circ}$. In addition, MDF of UT muscles in $0^{\circ}$ and $30^{\circ}$ of visual feedback was significantly greater than $60^{\circ}$. Conclusion: We concluded that $0^{\circ}$ visual feedback of head angle with using a mobile posture-aware system would be beneficial to prevent or decrease forward head posture during watching a smartphone. We also could recommend using of $30^{\circ}$ visual feedback in case of caring UT muscle fatigue primarily.

본 연구는 전 세계적으로 급증하는 한국어 학습 수요에 대응하여, 스마트폰만으로 정밀한 발음 교정이 가능한 시스템을 개발하는 것을 목적으로 한다. 기존의 2차원 평면 기반 음향 분석이 간과해 온 발화의 시공간적 안정성 (Stability)과 한국어 모음 변별의 핵심 기제인 입술 모양(lip rounding)을 통합적으로 분석한다. 연구 방법으로는 모바일 환경에서 선형 예측 부호화(LPC)를 통해 포먼트를 추출하고 커널 밀도 추정(KDE)을 적용하여 ‘3차원 포먼트 밀도’ 지형도 를 생성하며, 동시에 MediaPipe Face Mesh 기술로 입술의 개구도와 원순성을 정량화한다. 연구 결과, 개발된 시스템은 학습자가 생성하는 모음의 음향적 분산을 3차원 산맥 형태로 시각화하여 발음의 견고성을 인지하게 하였으며, 유사한 포먼 트 값을 가지더라도 입술 모양이 잘못된 경우를 실시간으로 탐지하여 교정 효율을 높였다. 본 연구는 ‘점’ 중심의 발음 평가를 ‘분포’ 중심으로 전환하고, 음성학과 컴퓨터 비전 기술을 융합하여 모바일 컴퓨터 기반 컴퓨터 보조 발음훈련 (computer-assisted pronunciation training; CAPT) 시스템의 새로운 표준을 제시하였다.

The rapid globalization of Korean culture has precipitated a surge in Korean language learners, yet effective pronunciation training remains a significant pedagogical bottleneck. Traditional Computer-Assisted Pronunciation Training (CAPT) systems typically rely on two-dimensional acoustic analysis or automatic speech recognition confidence scores. These methods fundamentally fail to capture the temporal stability of vowel production—treating pronunciation as a static point rather than a dynamic distribution—and neglect the critical articulatory role of lip kinematics. This research presents the development of a novel smartphone-based biofeedback system that integrates 3D Vowel Space Analysis with Real-Time Lip Contour Extraction. By employing Kernel Density Estimation (KDE) on accumulated formant trajectories, the system visualizes the "density" of a learner's pronunciation in three dimensions. Simultaneously, a computer vision pipeline utilizes the MediaPipe Face Mesh to extract facial landmarks, providing immediate visual feedback on lip rounding. This multimodal approach effectively decouples acoustic errors from articulatory misconfigurations.

Previous studies present the mixed results on online reputation mechanism. In this study, we have found that an approach based on Bayesian statistics can explain most results of previous studies which are conflicting with each others. With this model, we explain why negative ratings have more significant marginal impacts on sellers' reputation than positive ones do. Furthermore, we even show why the feedbacks with a few negative ratings may increase the value of the item and final prices by confirming buyers' prior beliefs on the sellers' reputation much more than those without negative ratings. Also, we explain why there are not many negative ratings. Even though some studies suggest this because of generosity of users, our model shows that the reason is that the existence of FS itself prevents bad sellers from participating to the market as a signal itself. Even further, we show how this extreme tendency of positive ratings gets even stronger as markets evolve. Finally, to validate our analytical results, we examine the previous studies and see what factors effect the outcomes of their analyses.

Background: We developed a machine vision technology program that tracks patients’ realtime breathing and automatically analyzes their breathing patterns. Materials and Methods: To evaluate its potential for clinical application, the image tracking performance and accuracy of the program were analyzed using a respiratory motion phantom. Changes in the stability and regularity of breathing were observed in healthy adult volunteers according to whether the breathing pattern mirrored the breathing guidance. Results and Discussion: Displacement within a few millimeters was observed in real-time with a clear resolution, and the image tracking ability was excellent. This result was consistent even in the sections where breathing patterns changed rapidly. In addition, the respiratory gating method that reflected the individual breathing patterns improved breathing stability and regularity in all volunteers. Conclusion: The findings of this study suggest that this technology can be used to set the appropriate window and the range of internal target volume by reflecting the patient’s breathing pattern during radiotherapy planning. However, further studies in clinical populations are required to validate this technology.

This study explores the effect of proprioceptive neuromuscular facilita-tion (PNF) exercises combined with a virtual reality (VR) motion capture system and concurrent feedback (CF) on early shoulder muscle activa-tion in healthy individuals. Thirty healthy participants sequentially per-formed three PNF diagonal 2 shoulder exercises: PNF alone, PNF with VR (PNF+VR), and PNF with VR and CF (PNF+VR+CF), with the latter two exercises presented in randomized order. Using wireless surface elec-tromyography and 3-dimensonal (3D) inertial measurement units, the activation of the upper trapezius (UT), lower trapezius (LT), infraspinatus (INF), and serratus anterior (SA) muscles, as well as shoulder flexion, abduction, and external rotation range of motion, were recorded during three exercises. INF and LT muscles were activated earlier in abduction and flexion during PNF+VR and PNF+VR+CF exercises than in PNF alone (P<0.05). These muscles were also activated during the early external rotation phase during PNF alone (P<0.05). No statistically significant differences were observed in UT and SA muscle activation among PNF alone, PNF+VR, and PNF+VR+CF exercises (P>0.05). Integrating VR gaming elements and CF into PNF shoulder exercises may promote ear-ly muscle activation, offering a promising approach for advanced reha-bilitation strategies that can potentially improve patient outcomes. VR motion capture systems have the potential to transform traditional ther-apeutic approaches by increasing enjoyment and engage

The problem of choosing the right product that will best fit the consumers’ tastes and preferences also extends even in electronic commerce. However, e-commerce was able to create a technological proxy for this social filtering process that is called online Recommender Systems (RS). RS act as personalized decision guides, aiding users in decisions on matters related to personal taste. It has the potential to support and improve the quality of the decisions consumers make when searching for and selecting products and services online. However, most previous research on RS has focused on the statistical accuracy of the algorithms driving the systems, with little emphasis on interface issues and the user’s perspective. This study identified transparency and feedback as some of the possible ways to effectively evaluate recommender systems based from the users’ perspective. Thus, the goal of this research wants to focus on examining and identifying the roles of transparency and feedback in recommender systems and how it affect the user’s attitude towards the system.

This study aimed to develop a sensory feedback system which could measure force and temperature for the user of myoelectric prosthetic hands. The Sensory measurement module consisted of a force sensing resistor to measure forces and non-contact infrared temperature sensor. These sensors were attached on the fingertips of the myoelectric prosthetic hand. The module was validated by using standard weights corresponding to external force and a Peltier module. Sensory transmission module consisted of four vibration motors. Eight vibration patterns were generated by combining motion of each vibration motor and were dependent on kinds and/or magnitude. The module was verified by using standard weigts and water at varying temperatures. There were correlations of force and temperature between the sensory measurement module and standard weight and water. Additionally, exact vibration patterns were generated, indicating the efficacy of the sensory feedback system for the myoelectric prosthetic hand.

As many sensors and actuators are used in many automated systems, various industrial networks are adopted for real-time distributed control. In order to take advantages of the networking, however, the network implementation should be carefully designed to satisfy real-time requirements considering network delays. This paper presents the implementation of a remote feedback control system via Profibus-DP net work for real-time distributed control More specifically, the effect of the network delay on the control performance evaluated on Profibus-DP testbed. Also, the traditional PID gain tuning methods are used to demonstrate the feasibility of the remote feedback control.

의사들은 학술연구와 환자진료에 관한 정보요구를 가지고 있고, 전자정보원을 이용하여 해결하고 있으나 정보원의 선택이나 검색에서 어려움을 겪는다. 본 연구는 의사들의 연구와 진료정보요구를 지원할 수 있는 검색시스템을 설계하고 개발하는 것이 목적이다. 이를 위해 의사들의 정보요구에 적합한 정보원 개발을 위해서 정보이용행태에 관하여 심층인터뷰를 하였고, 그 결과를 바탕으로 검색 시스템인 MediSearching을 개발하였다. 개발된 시스템은 효용성 검증과 제안을 얻기 위해 추가 인터뷰와 테스트검색을 통하여 피드백을 받았다. 의사들의 정보이용행태에 관한 인터뷰 결과 소속병원의 종류나 진료 과목에 따라서 차이를 보였다. 대학병원 의사들은 연구와 관련된 정보요구가 컸으며 학술지 논문을 주로 이용하고 있었으나, 전문병원이나 개인병원 의사들은 진료와 관련된 정보요구가 대부분으로 동료와의 대화로 해결하는 형태를 보였다. MediSearching은 연구와 진료정보요구를 동시에 지원할 수 있는 검색시스템으로 개발되었고, 인터페이스와 검색효용성이 검증되었다. 기존의 한국 의학 정보 사이트에서는 제공 정보의 종류나 검색방법에 따라 서비스 범위가 모두 달라 의사들은 동일한 검색 과정을 반복하는 불편함이 있었으나 이를 해결한 것으로 파악되었다. 또한 파일 형태나 언어와 상관없이 정보원의 추가가 가능하도록 한 점에서 유용한 것으로 평가되었다.

Physicians have information needs related to academic research and clinical practice but encounter difficulties seeking appropriate medical resources and information. The goal of this study is to develop a search system to support Korean physicians' information needs. To access sources to meet the identified need, in-depth interviews were conducted, and MediSearching, a new search system, was developed accordingly. To verify its usefulness and to obtain users' suggestions, interviews were conducted and feedback was solicited via collected test searches. The initial set of interviews indicated that physicians' information needs and search behaviors differed by type of hospital and clinical department. Physicians in university hospitals with a greater need for research materials searched for academic articles, whereas physicians in specialty hospitals or primary care clinics with a stronger need for materials related to clinical practice asked their colleagues for information. Consequently, MediSearching was designed to satisfy both groups' needs, and the test search provided useful search results compared with existing services. Korean physicians have previously had to repeat their search process on separate sites that provide different services according to type of information and search method. MediSearching may reduce this inconvenience and add documents in various formats and languages.

5G의 상용화 등 네트워킹 기술의 발전은 다양한 시스템들이 네트워크를 통해서 정보를 교환하고 제어할 수 있는 기반을 제공하고 있다. 또한, 네트워크에서 발생하는 많은 현상들은 정보의 지연과 관련되기 때문에 지연된 정보를 갖는 시스템의 제어의 중요성이 증가하고 있다. 본 논문에서는 최근들어 지연이 있을 때에 저복잡도 제어기 설계에 많이 활용되는 예측 제어를 도입하여, 지연된 다중 입력을 갖는 시스템에서 지연의 크기와 입력의 수에 상관없이 거의 일정한 복잡도를 갖는 예측 제어기를 제시한다. 또한, 출력 궤환 구조를 갖는 예측 제어기가 점근적 수렴이 보장됨을 증명한다. 모의 실험을 통해서 제안된 방식이 상태 벡터를 확장한 전통적인 방식이나, 다른 예측 기반 제어 방식에 비해 적은 복잡도를 가지면서 안정성을 보장하는 제어기 설계 성공이 높게 발생함을 확인하였다.

The evolution of networking technology such as commercialization of 5G systems provides foundation for information exchange and control of systems over the network. In addition, importance of controlling a system with delay is increasing significantly, since various phenomena in the network are associated with delay. In this paper, with a predictive control which has been studied for designing a controller with low complexity, we propose a novel predictive control for a system with multi-inputs such that it can keeps the complexity almost the same regardless of the number of inputs and degree of delay. The asymptotic stability of the proposed control with a static output feedback is also proved. The numerical simulation shows that the proposed method is superior in complexity and the performance of finding feasible controllers to the existing predictive control and a conventional method based on augmented states.

기계 학습과 인공 지능 기술의 발전으로 다양한 응용분야들이 가능해지고 있고, 이중에 추천 시스템은 이미 여러 업체들에서 영화 추천이나 상품 추천 등의 서비스에 적용하여 효과를 보고 있다. 이러한 서비스 중인 추천 시스템들의 대부분은 아이템의 내용을 분석하여 추천하거나 아니면 평점과 같은 직접적인 피드백에 기반하여 시스템을 학습하고 추천하고 있다. 하지만 많은 온라인 쇼핑몰 중에는 아이템의 내용을 분석하는 것이 어렵고, 직접적인 피드백 정보가 없거나 혹은 거의 없어 추천 시스템 구축이 어려운 경우가 많다. 이러한 경우에도 사용자의 상품 조회에 관한 로그 기록들은 어렵지 않게 확보할 수 있고, 로그 기록들만 가지고도 추천 서비스를 제공할 수 있다면 서비스의 질을 향상할 수 있을 것으로 기대된다. 본 논문에서는 사용자의 로그 기록으로부터 암묵적인 피드백인 상품 조회 정보를 추출하고, 암묵적인 피드백에 기반한 추천 시스템을 구현하고, 제안된 시스템은 온라인 반려동물 용품점에 적용하여 확인한다. 즉, 사용자들의 상품조회를 위한 클릭정보만을 활용하여 반려동물 용품 추천 시스템을 구축하여 서비스로 확인한다.

Due to the advances in machine learning and artificial intelligence technologies, many new services have become available. Among such services, recommendation systems have already been successfully applied to commercial services and made profits as in online shopping malls. Most recommendation algorithms in commercial services are based on content analysis or explicit feedback rates as in movie recommendations. However, many online shopping malls have difficulties in content analysis or are lacking explicit feedbacks on their items, which results in no recommendation system for their items. Even for such service systems, user log data is easily available, and if recommendations are possible with such log data, the quality of their service can be improved. In this paper, we extract implicit feedback like click information for items from log data and provide a recommendation system based on the implicit feedback. The proposed system is applied to a real in-service online shopping mall.

인공지능 기반 보안관제 시스템은 운영환경에서 발생할 수 있는 학습 데이터 오류, 신규 공격 이벤트 발생으로 인한 오탐 증가 등 문제를 해결하기 위해 피드백 기능이 연구되고 있다. 그러나 한정된 관제 인력의 피드백 수행 방식은 모델 개선에 오랜 시간이 걸리고 숙련되지 않은 관제 인력의 피드백은 오히려 모델 성능 저하의 원인이 될 수 있다. 본 논문에서는 관제 인력 한계 극복, 신규 오탐 개선, 빠른 모델 성능 향상을 위한 능동형 보안관제 모델 개선 프로세스를 제안하였다. 운영 중 예측된 유사 이벤트를 군집화 하고, 피드백이 우선적으로 필요한 군집을 계산하여 운영자에게 대표 이벤트 설명이 가능한 인공지능(eXplainable AI) 기반 시각화도 함께 제시하였다. 수신된 대표 피드백은 동일 군집과 다른 데이터를 계산하여 제외하고 피드백 전파 학습 데이터를 생성한다. 준비된 학습 데이터는 초기 모델과 함께 점진적 학습을 통해 모델을 생성함으로써 성능을 향상시키는 프로세스이다. 제안 프로세스의 실효성 검증을 위해 웹 어플리케이션 방화벽 데이터셋 PKDD2007과 CSIC2012를 선택하여 3개의 시나리오를 통해 실험을 진행하였다. 실험 결과 제안된 프로세스는 피드백을 주지 않았거나 소수 운영자 피드백을 적용한 모델 성능에 비해 모든 지표에서 약 30% 이상의 성능 향상을 확인하였다.

In the field of SIEM(Security information and event management), many studies try to use a feedback system to solve lack of completeness of training data and false positives of new attack events that occur in the actual operation. However, the current feedback system requires too much human inputs to improve the running model and even so, those feedback from inexperienced analysts can affect the model performance negatively. Therefore, we propose "active model improving feedback technology" to solve the shortage of security analyst manpower, increasing false positive rates and degrading model performance. First, we cluster similar predicted events during the operation, calculate feedback priorities for those clusters and select and provide representative events from those highly prioritized clusters using XAI (eXplainable AI)-based event visualization. Once these events are feedbacked, we exclude less analogous events and then propagate the feedback throughout the clusters. Finally, these events are incrementally trained by an existing model. To verify the effectiveness of our proposal, we compared three distinct scenarios using PKDD2007 and CSIC2012. As a result, our proposal confirmed a 30% higher performance in all indicators compared to that of the model with no feedback and the current feedback system.

This paper presents a position control strategy for a pump-controlled electro-hydrostatic actuator (EHA) using feedforward control with disturbance compensation. As the disturbance observer is used to estimate nonlinear dynamics of EHA, which has valve-opening conditionals, as well as external disturbances, an additional feedforward control is adopted to achieve rapid response. The effectiveness of the proposed control strategy is verified through experiment using an EHA test bench. The proposed controller shows better tracking performance compared with a conventional PID controller.

In the manufacture of flat panel display devices, there is a strong demand for contactless glass panel handling devices that can manipulate a glass panel without contaminating or damaging it. To fulfill this requirement, an electrostatic suspension device far glass panels where the glass panel is supported by electrostatic forces without any mechanical contact is proposed. To implement the system with low cost and compactness, switched-voltage control scheme that is based on the relay feedback control is utilized. Relay feedback control method deploys only a single high-voltage power supply that can deliver a DC voltage of positive and/or negative polarity and thus high voltage amplifiers that are costly and bulky are not needed any more. It is shown that despite the inherent limit cycle property of the relay feedback based control, an excellent performance in vibration suppression is attained due to the presence of a relatively large squeeze film damping originating from the electrodes and levitated object. Using this scheme, a $100{\times}100mm^2$ glass panel was levitated stably with airgap variation decreasing down to $1\;{\mu}m$ at an airgap of $100\;{\mu}m$.