서비스 보안 수준을 고려한 서버 클러스터링 기반 침입감내시스템 설계

권현,이용재,윤현수 육군사관학교 화랑대연구소 2016 한국군사학논집 Vol.72 No.2

Internet is an open space where a number of computer systems are connected to one another. Unfortunately, as systems provide many functionalities with users, they have vulnerabilities that can be used by malicious users who try to intrude into a system. Although such malicious activities by either internal or external adversaries can be defended by conventional security systems such as Intrusion Detection and Prevention System (IDPS), it is not always possible to defend a target system against the attacks completely. For this reason, Intrusion Tolerance System (ITS) has been proposed to maintain service provision even under threatening environments, where some attacks succeed in part. In this paper, we propose an ITS based on the server clustering scheme where servers are grouped into a cluster according to the security level because each service requires a different security level. The proposed scheme allocates spare servers to a cluster, called a security cluster, since more resources should be ready to be assigned to the security cluster to maintain a higher security level. By this way, the proposed system can maintain the performance of services to clients. 최근 적 사이버전 수행으로 인해서 컴퓨터 시스템 내에 취약점이 증대되고 있고 악의적인 해커들을 통해서 취약점을 이용한 공격이 이뤄지고 있다. 비록 기존 방어솔루션인 IDS나 IPS를 통해서 내외부적 공격을 방어하고 있지만, 제로데이와 같이 알려지지 않은 공격 등으로 부터 시스템을 완벽히 방어하기에는 불가능하다. 따라서 침입감내시스템은 적의 침입을 어느정도 허용하더라도 중요서비스를 지키고 정상적인 서비스를 제공하는 데에 목표가 있다. 이 논문에서는 서버 클러스터링을 통한 침입감내시스템을 제안하였다. 서비스별 중요도에 따라서 Security cluster와 Normal cluster로 구분하여서 보안성이 중요시 되는 서비스를 보호하면서도 중요 서비스의 최소 서비스 보장을 유지할 수 있는 침입감내시스템을 제안하였다. Cloudsim 시뮬레이터를 통해서 보안성 수준과 요구되는 가상화머신의 수를 기존 침입감내시스템인 SCIT(Self Cleansing Intrusion Tolerance)와 비교분석하였다.

공급사슬 안정화를 위한 화물운송산업의 안전운임제 도입 효과분석 - 군집별 차별적 효과를 중심으로 -

백두주,정현일 한국산업노동학회 2021 산업노동연구 Vol.27 No.3

The purpose of this study is to analyze the effects of the introduction of the safe rates system in the cargo transportation industry, which has implemented since January 2020 in South Korea and to explore the policy implications for low feeling groups. To this end, factors were classified by domain through factor analysis on the feeling effect of the safe fare system, and cluster types were classified according to effect of each factor through hierarchical clustering and K-means clustering. In addition, to confirm the characteristics of each cluster, ANOVA was performed on demographic characteristics, working conditions, job satisfaction, and system compliance rate and cross-analysis was performed on business types. As a result of the analysis, the feeling effect of the safe rates system was classified into factors for each of four areas: work-life balance, quality of life, safety, and industrial structure. Three clusters were derived from cluster analysis: The first cluster was ‘partially feeling type’, which showed a positive effect only on safety and industrial structure, the second cluster was no feeling type, which showed negative effects in all factors, and the third Cluster is a strong feeling type that showed a positive effect in all factors. These clusters showed significant differences in demographic characteristics, working conditions, job satisfaction, system compliance rate, and business type characteristics. Policy implications include supplementing the announcement of additional provisions of the safe fare system and supporting collective agreement for workers in the BCT long-distance transportation type, which has the highest proportion of no feeling types, activation of Safe rate report center, improvement of sunset restriction for increasing sustainability on safe rates system. 이 연구의 목적은 2020년 1월부터 시행 중인 화물운송산업의 안전운임제 도입효과를 군집별로 분석하고 긍정적 체감수준이 낮은 집단을 위한 정책적 함의를 탐색하는 것이다. 이를 위해 안전운임제 체감효과에 대한 요인분석을 통해 영역별 요인들을 구분하고, 계층적 군집분석과 K-평균 군집분석으로 요인별 효과에 따른 군집유형을 구분했다. 그리고 군집별 특성을 확인하기 위해 인구학적 특성, 노동조건, 직업만족도, 제도준수율에 대한 분산분석과 업태별 교차분석 수행했다. 분석 결과, 안전운임제의 체감효과는 일과 삶의 균형, 삶의 질, 안전, 산업구조 4개 영역별 요인으로 분류되었다. 군집분석으로 3개의 군집이 도출되었는데 첫 번째 군집은 안전운임의 안전, 산업구조에서만 긍정적 효과를 보인 ‘부분 체감형’, 두 번째 군집은 모든 요인에서 부정적 효과를 보인 ‘미체감형’, 세 번째 군집은 모든 요인에서 긍정적 효과를 보인 ‘강한 체감형’이다. 이들 군집은 인구학적 특성, 노동조건, 직업만족도, 제도준수율 및 업태 특성에서 유의미한 차이를 보였다. 정책적 함의로는 미체감형 비중이 가장 높은 BCT 장거리 운송형태 종사자들을 위한 안전운임제 부대조항 고시 보완 및 업태-운송형태별 집단협약 지원, 제도준수율을 높이기 위한 안전운임신고센터 활성화, 마지막으로 제도의 지속성을 높이기 위한 일몰제도 폐지를 제시했다.

대규모 단체급식 데이터를 활용한 음식 메뉴 군집화와 추천 시스템 성능 비교

방병권,김민용 대한경영정보학회 2021 경영과 정보연구 Vol.40 No.2

데이터의 양이 기하급수적으로 증가하고 이를 처리할 수 있는 분석 방법이 발전함에 따라 개인의 선호를 반영한 제품과 서비스를 제안하기 위해 다양한 분야에서 추천 시스템이 활용되고 있다. 추천 시스템을 통하 여 소비자의 선호를 예측할 수 있게 되면 제품과 서비스에 대한 수요예측이 보다 정확해지고 이는 재고관리 등에 있어서 자원을 보다 효율적으로 운영할 수 있다. 대규모의 소비자를 상대로 선호를 파악하고 이를 수요 예측에 반영하는 방법으로 추천 시스템을 활용할 수 있으나, 음식 메뉴의 추천을 위해 선택이 제한된 상황에 서의 소비자 선택 정보를 활용하여 군집화를 시도하거나 추천 시스템의 성능을 비교한 연구는 드물다. 본 연구에서는 대규모 단체급식에서 발생한 음식 메뉴에 대한 소비자의 선택정보를 활용하여 음식 메뉴를 군집화하고 추천 시스템 알고리즘 간의 성능을 비교하였다. 단체 급식에는 다양한 선호를 가지고 있는 소비자 들을 만족시키기 위해 다양한 음식 메뉴를 제공할 수 없는 한계가 있다. 소비자의 선호를 보다 정확하게 반영 하게 되면 제한된 음식 메뉴로도 소비자의 만족도를 향상시킬 수 있고 식사를 제공하는 입장에서는 잔반을 최 소화하고 원재료의 적정 재고 운영을 통해 자원의 효율적인 운영이라는 목적을 달성할 수 있다. 따라서 본 연 구에서는 대규모 단체급식에서 발생한 음식 메뉴에 대한 소비자의 선택과 비선택의 정보를 활용하여 음식 메 뉴의 군집화를 시도하고, 나이브베이즈 분류기를 통해 각 군집의 특성을 제시하였다. 또한 2~3개의 선택 가능한 메뉴 중에서 소비자가 선택을 하는 음식 메뉴에 대한 선택정보만을 반영하였을 경우와 비선택의 경쟁 정보를 추가하여 반영한 두 가지 경우에 대한 추천 시스템 알고리즘 간의 성능 차이를 비교하였다. 단체급식 데이터를 활용하여 군집 분석을 한 결과 음식 메뉴는 8개의 군집으로 나눌 수 있었으며 각 군집 의 특성은 나이브베이즈 분류를 통해 소비자의 인구통계 정보 및 음식 메뉴의 특성으로 설명이 가능한 것으 로 나타났다. 음식 메뉴의 추천 시스템 성능은 대중적인 메뉴를 추천하는 경우가 예측 성능이 가장 우수한 것으로 나타났으며, 소비자가 선택한 아이템에 대한 정보뿐만 아니라 선택되지 않은 정보를 반영하면 추천 시스템 성능이 향상되는 것으로 나타났다. 이러한 연구는 음식 메뉴의 추천에 있어서 군집화 정보를 활용하고 선택의 기회가 있는 경우에는 비선택 의 경쟁 정보까지 반영하여 추천 시스템을 구성함으로써 추천 성능을 향상시킬 수 있다는 것을 제시하고 있 다. 대규모 단체급식에서 소비자의 선호를 바탕으로 군집화를 하고 추천의 정확도를 높이는 연구는 자원의 효율적 운영을 위한 수요예측의 기초자료로 활용이 될 수 있다. As the amount of data increases exponentially and an analysis method capable of processing it develops, recommendation systems are being used in various fields to propose products and services that reflect individual preferences. When consumers' preferences can be predicted through the recommendation system, demand forecasts for products and services become more accurate, and resources can be operated more efficiently in inventory management, etc. The recommendation system can be used as a method of identifying preferences for large-scale consumers and reflecting them in demand forecasting. But there are few studies that attempts to cluster by using consumer selection information in situations where selection is limited for recommendation of meal menus or performance of the recommendation system. In this study, meal menus were clustered using the consumer's selection information for meal menus generated in large-scale company cafes, and the performance of recommendation system algorithms was compared. Large-scale company cafes have a limit in that they cannot provide a variety of meal menus to satisfy consumers with various preferences. If consumers' preferences are more accurately reflected, consumers' satisfaction can be improved even with limited meal menus, and from the standpoint of providing meals, the goal of efficient operation of resources can be achieved by minimizing leftovers and operating an appropriate inventory of raw materials. Therefore, in this study, we attempted to cluster meal menus by using the information of consumers' choice and non-selection on meal menus that occurred in large-scale company cafes and presented the characteristics of each cluster through the Naive Bayes classifier. In addition, performance differences between the recommendation system algorithms were compared in the case of reflecting only the selection information for the meal menu selected by the consumer from among 2 to 3 selectable menus, and the two cases reflecting by adding competition information of non-selection. As a result of cluster analysis using group meal data, the meal menu could be divided into 8 clusters, and the characteristics of each cluster could be explained by the consumer's demographic information and the characteristics of the meal menu through the Naive Bayes classification. As for the performance of the meal menu recommendation system, the case of recommending the popular menu showed the best prediction performance, and it was found that the performance of the recommendation system was improved by reflecting not only the information on the item selected by the consumer but also the information not selected. This study suggests that the recommendation performance can be improved by constructing a recommendation system by utilizing clustering information in recommending meal menus and reflecting competition information of non-selection when there is an opportunity for selection. In large-scale group catering, research on clustering based on consumer preferences and increasing the accuracy of recommendations can be used as basic data for demand forecasting for efficient management of resources.

추천 시스템의 성능 안정성을 위한 예측적 군집화 기반 협업 필터링 기법

이오준(O-Joun Lee),유은순(Eun-Soon You) 한국지능정보시스템학회 2015 지능정보연구 Vol.21 No.1

With the explosive growth in the volume of information, Internet users are experiencing considerable difficulties in obtaining necessary information online. Against this backdrop, ever-greater importance is being placed on a recommender system that provides information catered to user preferences and tastes in an attempt to address issues associated with information overload. To this end, a number of techniques have been proposed, including content-based filtering (CBF), demographic filtering (DF) and collaborative filtering (CF). Among them, CBF and DF require external information and thus cannot be applied to a variety of domains. CF, on the other hand, is widely used since it is relatively free from the domain constraint. The CF technique is broadly classified into memory-based CF, model-based CF and hybrid CF. Model-based CF addresses the drawbacks of CF by considering the Bayesian model, clustering model or dependency network model. This filtering technique not only improves the sparsity and scalability issues but also boosts predictive performance. However, it involves expensive model-building and results in a tradeoff between performance and scalability. Such tradeoff is attributed to reduced coverage, which is a type of sparsity issues. In addition, expensive model-building may lead to performance instability since changes in the domain environment cannot be immediately incorporated into the model due to high costs involved. Cumulative changes in the domain environment that have failed to be reflected eventually undermine system performance. This study incorporates the Markov model of transition probabilities and the concept of fuzzy clustering with CBCF to propose predictive clustering-based CF (PCCF) that solves the issues of reduced coverage and of unstable performance. The method improves performance instability by tracking the changes in user preferences and bridging the gap between the static model and dynamic users. Furthermore, the issue of reduced coverage also improves by expanding the coverage based on transition probabilities and clustering probabilities. The proposed method consists of four processes. First, user preferences are normalized in preference clustering. Second, changes in user preferences are detected from review score entries during preference transition detection. Third, user propensities are normalized using patterns of changes (propensities) in user preferences in propensity clustering. Lastly, the preference prediction model is developed to predict user preferences for items during preference prediction. The proposed method has been validated by testing the robustness of performance instability and scalability-performance tradeoff. The initial test compared and analyzed the performance of individual recommender systems each enabled by IBCF, CBCF, ICFEC and PCCF under an environment where data sparsity had been minimized. The following test adjusted the optimal number of clusters in CBCF, ICFEC and PCCF for a comparative analysis of subsequent changes in the system performance. The test results revealed that the suggested method produced insignificant improvement in performance in comparison with the existing techniques. In addition, it failed to achieve significant improvement in the standard deviation that indicates the degree of data fluctuation. Notwithstanding, it resulted in marked improvement over the existing techniques in terms of range that indicates the level of performance fluctuation. The level of performance fluctuation before and after the model generation improved by 51.31% in the initial test. Then in the following test, there has been 36.05% improvement in the level of performance fluctuation driven by the changes in the number of clusters. This signifies that the proposed method, despite the slight performance improvement, clearly offers better performance stability compared to the existing techniques. Further research on this study will be direct

Cluster-based multicore real-time mixed-criticality scheduling

Ali, A.,Kim, K.H. Elsevier 2017 JOURNAL OF SYSTEMS ARCHITECTURE - Vol.79 No.-

Cluster-based scheduling is recently gaining importance to be applied to mixed-criticality real-time systems on multicore processors platform. In this approach, the cores are grouped into clusters, and tasks that are partitioned among different clusters are scheduled by global scheduler in each cluster. This research work introduces a new cluster-based task allocation scheme for the mixed-criticality real-time task sets on multicore processors. For task allocation, smaller clusters sizes (sub-clusters) are used for mixed-criticality tasks in low criticality mode, while relatively larger cluster sizes are used for high criticality tasks in high criticality mode. In this research paper, the mixed-criticality task set is allocated to clusters using worst-fit heuristic. The tasks from each cluster are also allocated to its sub-clusters, using the same worst-fit heuristic. A fixed-priority response time analysis approach based on Audsley's approach is used for the schedulability analysis of tasks in each cluster and sub-cluster. If the high criticality job is not completed after its worst case execution time in low mode, then the system is switched to high criticality mode. After mode switch, all the low criticalities tasks are discarded and only high criticality tasks are further executed in high criticality mode. Simulation results indicate that the percentage of schedulable task sets significantly increases under cluster scheduling as compared to partitioned and global mixed-criticality scheduling schemes.

Multi-Feature Clustering을 이용한 강인한 내용 기반 음악 장르 분류 시스템에 관한 연구

윤원중,이강규,박규식,Yoon Won-Jung,Lee Kang-Kyu,Park Kyu-Sik 대한전자공학회 2005 電子工學會論文誌-SP (Signal processing) Vol.42 No.3

In this paper, we propose a new robust content-based musical genre classification algorithm using multi-feature clustering(MFC) method. In contrast to previous works, this paper focuses on two practical issues of the system dependency problem on different input query patterns(or portions) and input query lengths which causes serious uncertainty of the system performance. In order to solve these problems, a new approach called multi-feature clustering(MFC) based on k-means clustering is proposed. To verify the performance of the proposed method, several excerpts with variable duration were extracted from every other position in a queried music file. Effectiveness of the system with MFC and without MFC is compared in terms of the classification accuracy. It is demonstrated that the use of MFC significantly improves the system stability of musical genre classification performance with higher accuracy rate. 본 논문에서는 multi-feature clustering(MFC) 방법을 이용한 강인한 내용 기반 음악 장르 분류 알고리즘을 제안한다. 기존 연구와 비교하여 본 논문에서는 입력 질의 패턴(또는 구간)과 입력 질의 길이의 변화에 따라 나타나는 불안정한 시스템 성능을 개선하는데 노력하였고, k-means clustering 기법에 기반한 multi-feature clustering(MFC)이라는 새로운 알고리즘을 제안하였다. 제안된 시스템의 성능을 검증하기 위해 질의 음악 파일의 서로 다른 여러 구간에서 질의 길이를 다변화하여 음악 특징 계수를 추출하였고, MFC 방법을 사용한 시스템과 MFC 방법을 사용하지 않은 시스템에 대한 장르 분류 성공률을 비교하여 제안 알고리즘의 성능을 비교${\cdot}$분석하였다. 모의실험 결과 MFC 방법을 사용한 시스템의 장르 분류 성공률이 높게 나타났고, 시스템의 안정성 역시 높게 나타났다.

K-Means Clustering with Content Based Doctor Recommendation for Cancer

Rethina kumar,Gopinath Ganapathy,Jeong-Jin Kang 국제문화기술진흥원 2020 International Journal of Advanced Culture Technolo Vol.8 No.4

Recommendation Systems is the top requirements for many people and researchers for the need required by them with the proper suggestion with their personal indeed, sorting and suggesting doctor to the patient. Most of the rating prediction in recommendation systems are based on patient’s feedback with their information regarding their treatment. Patient’s preferences will be based on the historical behaviour of similar patients. The similarity between the patients is generally measured by the patient’s feedback with the information about the doctor with the treatment methods with their success rate. This paper presents a new method of predicting Top Ranked Doctor’s in recommendation systems. The proposed Recommendation system starts by identifying the similar doctor based on the patients’ health requirements and cluster them using K-Means Efficient Clustering. Our proposed K-Means Clustering with Content Based Doctor Recommendation for Cancer (KMC-CBD) helps users to find an optimal solution. The core component of KMC-CBD Recommended system suggests patients with top recommended doctors similar to the other patients who already treated with that doctor and supports the choice of the doctor and the hospital for the patient requirements and their health condition. The recommendation System first computes K-Means Clustering is an unsupervised learning among Doctors according to their profile and list the Doctors according to their Medical profile. Then the Content based doctor recommendation System generates a Top rated list of doctors for the given patient profile by exploiting health data shared by the crowd internet community. Patients can find the most similar patients, so that they can analyze how they are treated for the similar diseases, and they can send and receive suggestions to solve their health issues. In order to the improve Recommendation system efficiency, the patient can express their health information by a natural-language sentence. The Recommendation system analyze and identifies the most relevant medical area for that specific case and uses this information for the recommendation task. Provided by users as well as the recommended system to suggest the right doctors for a specific health problem. Our proposed system is implemented in Python with necessary functions and dataset.

클러스터링 기법을 이용한 소프트웨어 요구사항 적합성 검증도구 개발

박헌우,최재훈,황석근,박문식,노상욱,정기현,최경희 한국차세대컴퓨팅학회 2021 한국차세대컴퓨팅학회 논문지 Vol.17 No.6

소프트웨어의 요구사항을 작성하면서 발생하는 모호성과 부적절성을 제거하기 위하여 요구사항에 대한 전문용어를 사용할 필요가 있다. 본 논문에서는 철도차량 시스템 소프트웨어 요구사항 분석을 통하여 요구사항 작성에 참조할 수 있는 전문용어를 정립하며, 이를 기반으로 올바르게 작성된 요구사항 문장들에 대한 템플릿을 도출한다. 다양한 템플릿 문장은 요구사항 적합성을 점검하기 위한 비지도 기계학습용 데이터로 사용된다. 군집화 모델을 활용하여 요 구사항 템플릿을 군집화 하며, 새로운 요구사항의 적합성 여부를 군집 기반으로 판별하는 요구사항 적합성 검증도구 를 개발한다. 특정한 요구사항이 군집화 모델에 의하여 생성된 군집의 템플릿과 정확하게 일치하면 적합한 요구사항 으로 판정한다. 반면에 요구사항 적합성에 실패하면, k-평균 군집화 알고리즘에 의하여 가장 유사한 템플릿을 자동 으로 추천한다. 클러스터링 기법에 의하여 군집화된 템플릿과 새로운 요구사항에 대한 적합성 여부를 검증하기 위한 실험을 수행하였다. 첫 번째는 군집화 기반의 요구사항 추천이 정상적으로 동작하는가에 대한 실험이었다. 새로운 요구사항이 입력되면 군집으로 분류되었으며, 특정한 군집내의 동일한 형태소의 배열을 가진 템플릿으로 정확하게 추천됨을 확인하였다. 두 번째 실험에서는 기존의 템플릿 데이터베이스에 없는 요구사항 문장을 입력한 경우에 입력 한 요구사항과 가장 유사한 템플릿을 추천하는 것을 확인하였다. It is necessary for users to utilize a set of standard technical jargon for the soundness of requirements specification, while reducing its ambiguity and improperness as much as possible. Through the analysis of requirements specification, a standard technical Korean(STK) has been established in the domain of the railway vehicle system. Based upon STK, this paper also derives a type of templates for the guideline of requirements specification. A variety of templates in the railway vehicle system domain are fed into the input of unsupervised machine learning algorithms as training instances. The clustering models in an unsupervised way classify the templates into a several clusters. Our system that verifies the soundness of requirements specification has been developed in the basis of clusters. The system we have developed makes a new sentence of specification belonged into one of groups, and then, if it finds the identical template in a specific cluster, it confirms the new specification as a sound one. Otherwise, it autonomically returns the most similar template recommended by k-means clustering algorithm. We have tested our system to verify both cases. In the experiment, no matter what a new sentence exists in the knowledge base of templates, it turns out that our robust system confirms whether or not the new sentence is correctly written, based upon the clusters, and further provides the most similar template as being updated for the correct specification.

Comparisons on Clustering Methods: Use of LMS Log Variables on Academic Courses

( Il-hyun Jo ),( Yeonjeong Park ),( Jongwoo Song ) 한국교육공학회 2017 Educational Technology International Vol.18 No.2

Academic analytics guides university decision-makers to assign limited resources more effectively. Especially, diverse academic courses clustered by the usage patterns and levels on Learning Management System(LMS) help understanding instructors’ pedagogical approach and the integration level of technologies. Further, the clustering results can contribute deciding proper range and levels of financial and technical supports. However, in spite of diverse analytic methodologies, clustering analysis methods often provide different results. The purpose of this study is to present implications by using three different clustering analysis including Gaussian Mixture Model, K-Means clustering, and Hierarchical clustering. As a case, we have clustered academic courses based on the usage levels and patterns of LMS in higher education using those three clustering techniques. In this study, 2,639 courses opened during 2013 fall semester in a large private university located in South Korea were analyzed with 13 observation variables that represent the characteristics of academic courses. The results of analysis show that the strengths and weakness of each clustering analysis and suggest that academic leaders and university staff should look into the usage levels and patterns of LMS with more elaborated view and take an integrated approach with different analytic methods for their strategic decision on development of LMS.

음주에 대한 양가감정과 메타인지가 대학생의 음주행동에 미치는 영향

이새롬,조성근,이장한 한국임상심리학회 2012 Korean Journal of Clinical Psychology Vol.31 No.4

Risk-taking behaviors result from the imbalance between affective Hot system and rational Cool system. This imbalance is affected by ambivalence and metacognition. The purpose of this study was to investigate the effect of ambivalence and metacognition on drinking behaviors as risk-taking behaviors. In order to create naturalistic groups based on actual data, we used cluster analysis. A four-cluster solution was selected for representation of data. Cluster 1 represented high ambivalence and high metacognition, Cluster 2 represented low ambivalence and high metacognition, Cluster 3 represented low ambivalence and low metacognition, and Cluster 4 represented high ambivalence and low metacognition. In addition, the analysis examined differences in alcohol use within groups. Cluster 1 had a lower frequency of drinking, a smaller amount of drinking, a lower level of obsessive alcohol thought, and a superior stage of change, compared with the other clusters. Results of correlation analyses indicated that both ambivalence and metacognition were related to drinking behaviors. These results suggest that the balance between Hot system and Cool system is essential for healthy drinking. Therefore, understanding the balance between the systems may be needed in order to resolve drinking problems.