• Journal of Korea Multimedia Society
• /
• v.13 no.6
• /
• pp.841-848
• /
• 2010

Conventional filters using email header and body information equally judge whether an incoming email is spam or not. However this is unrealistic in everyday life because each person has different criteria to judge what is spam or not. To resolve this problem, we consider user preference information as well as email category information derived from the email content. In this paper, we have developed a personalized anti-spam system using ontologies constructed from rules derived in a data mining process. The reason why traditional content-based filters are not applicable to the proposed experimental situation is described. In also, several experiments constructing classifiers to decide email category and comparing classification rule learners are performed. Especially, an ID3 decision tree algorithm improved the overall accuracy around 17% compared to a conventional SVM text miner on the decision of email category. Some discussions about the axioms generated from the experimental dataset are given too.

• Journal of KIISE:Software and Applications
• /
• v.28 no.1
• /
• pp.64-72
• /
• 2001

결정트리 생성은 일련의 특징값으로 기술된 사례들로부터 분류 지식을 추출하는 학습 방법중의 하나이다. 현장에서 수집되는 사례들은 관측 오류, 주관적인 판단, 불확실성 등으로 인해서 애매하게 주어지는 경우가 많다. 퍼지숫자나 구간값을 사용함으로써 이러한 애매한 데이타의 수치 속성은 쉽게 표현될 수 있다. 이 논문에서는 수치 속성은 보통값 뿐마아니라 퍼지숫자나 구간값을 갖을 수 있고, 비수치 속서은 보통값을 가지며, 데이터의 클래스는 확신도를 기자는 학습 데이터들로 부터, 분류 규칙을 마이닝하기 위한 퍼지 결정트리 생성 방법을 제안한다. 또한 제안한 방법에 의해 생성된 퍼지 결정트리를 사용하여, 새로운 데이터에 대한 클래스를 결정하는 추론 방법을 소개한다. 한편, 제안된 방법의 유용성을 보이기 위해 수행한 실험의 결과를 보인다.

• Journal of KIISE:Software and Applications
• /
• v.26 no.3
• /
• pp.343-352
• /
• 1999

본 논문에서는 대용량 데이터베이스에서 효율적인 연관 규칙 탐사에 대한 알고리즘을 제안하였다. 제안하는 알고리즘은 복합 해쉬 트리를 사용하여 해쉬 트리 탐색 비용과 데이터베이스 스캔 비용을 동시에 줄임으로서 성능을 향상시켰다. 복합 해쉬 트리는 같은 크기의 항목집합들 대신에 크기가 다른 여러 항목집합을 하나의 해쉬 트리로 구성한다. 복합 해쉬 트리의 유용성을 보이기 위하여 제안한 알고리즘은 잘 알려져 있는 Apriori, DHP 방밥과 수행 시간 측면에서 성능 비교를 하였다. 그 결과 대부분의 최소 지지도에서제안한 알고리즘이 Apriori, DHP 방법보다 우수하게 나타났으며, 최소 지지도가 0.5% 이하인 경우 DHP 방법에 비하여 약 30%의 이득 향상이 있었다.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2000.04a
• /
• pp.271-282
• /
• 2000

본 연구의 목적은 데이터마이닝 기법을 임상적으로 중요한 위치를 차지하고 있는 고혈압 환자의 특성과 치료에 따른 예후를 예측할 수 있는 지식을 발굴하고 이의 임상적용의 타당성을 검증하여 의사결정지원시스템을 개발하고 이의 유용성을 평가하는데 있다. 이에 연세대학교 의과대학 부속 세브란스 병원의 환자를 대상으로 로지스틱 회귀분석을 이용하여 혈압조절상의 위험요인의 규명하고, 의사결정나무분석을 통해 치료약제별 혈압조절군과 비조절군의 특성을 도출하고 각 대상군을 결정짓는 규칙을 생성하였으며, 이를 활용한 의사결정지원시스템의 개발 및c 평가를 시행하였다. 그 결과 기존 임상이론만을 활용한 시스템의 처방에 의한 혈압조절군보다 데이터마이닝 기법을 활용한 시스템의 처방에 의한 혈압조절군의 비율이 전체적으로 더 높게 나타남을 알 수 있었다. 본 연구의 결과는 우리나라 현실에 부합되는 고혈압 진료지침을 개발하고 적용, 평가하는데 기여할 수 있을 것으로 판단되며, 이와 같은 의사결정지원 시스템을 운영을 통해 실제 임상 진료에 적용해 봄으로써 그 효과와 실증적 가치를 창출할 수 있을 것이다.

• Journal of the Korean Data and Information Science Society
• /
• v.18 no.1
• /
• pp.1-10
• /
• 2007

In this paper, we focus on churn analysis for the first successful candidates in the entrance examination on 2006 year using Clementine, data mining tool. The goal of this study is to apply decision tree including C5.0 and CART algorithms, neural network and logistic regression techniques to predict a successful candidate churn. And we analyze the churning and nochurning successful candidates and why the successful candidates churn and which successful candidates are most likely to churn in the future using data from entrance examination data of K university on 2006 year.

• Asia pacific journal of information systems
• /
• v.16 no.2
• /
• pp.69-84
• /
• 2006

Importance of Today's diffusion of integrated hospital information systems is that various and huge amount of data is being accumulated in their database systems. Many researchers have studied utilizing such hospital data. While most researches were conducted mainly for medical diagnosis, there have been insufficient studies to develop medical care cost prediction model, especially using machine learning techniques. In this research, therefore, we built a medical care cost prediction model for cancer patients using CBR (Case-Based Reasoning), one of the machine learning techniques. Its performance was compared with those of Neural Networks and Decision Tree models. As a result of the experiment, the CBR prediction model was shown to be the best in general with respect to error rate and linearity between real values and predicted values. It is believed that the medical care cost prediction model can be utilized for the effective management of limited resources in hospitals.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.265-270
• /
• 2006

In an organization where major business operations are geared by business process management system(BPMS), routine tasks are processed according to the predefined business processes. However, most business operations are subject to some sort of exceptions, and the exceptional situations require update of the existing business process model, or a new business process model has to be defined to handle the exceptions. This paper proposes a system architecture that deploys business process repository as the media for storage and retrieval of the various business process models developed for exception handling. Well defined situation variables and decision variables play the key role for efficient storage and retrieval of the business process models developed for exception handling. The data mining technique C5.0 was used to build the optimum path for the process repository search tree.

• Journal of the Korea Society of Computer and Information
• /
• v.10 no.2 s.34
• /
• pp.215-221
• /
• 2005

TRecently, there has been enormous growth in the amount of commercial and scientific data, such as protein sequences, retail transactions, and web-logs. Such datasets consist of sequence data that have an inherent sequential nature. In this Paper, we study how to classify these sequence datasets. There are several kinds techniques for data classification such as decision tree induction, Bayesian classification and K-NN etc. In our approach, we use a K-NN algorithm for classifying sequences. In addition, we propose a new similarity measure to compute the similarity between two sequences and an efficient method for measuring similarity.

• Proceedings of the KSRS Conference
• /
• 2008.10a
• /
• pp.406-409
• /
• 2008

Clustering data streams has an importance over many applications like sensor networks. Existing hierarchical methods follow a semi fuzzy clustering that yields duplicate clusters. In order to solve the problems, we propose an extended online divisive agglomerative clustering on data streams. It builds a tree-like top-down hierarchy of clusters that evolves with data streams using geometric time frame for snapshots. It is an enhancement of the Online Divisive Agglomerative Clustering (ODAC) with a pruning strategy to avoid duplicate clusters. Our main features are providing update time and memory space which is independent of the number of examples on data streams. It can be utilized for clustering sensor data and network monitoring as well as web click streams.

• Journal of Computing Science and Engineering
• /
• v.8 no.2
• /
• pp.87-93
• /
• 2014

Coronary artery disease (CAD) is currently a prevalent disease from which many people suffer. Early detection and treatment could reduce the risk of heart attack. Currently, the golden standard for the diagnosis of CAD is angiography, which is an invasive procedure. In this article, we propose an algorithm that uses data mining techniques, a fuzzy expert system, and the imperialist competitive algorithm (ICA), to make CAD diagnosis by a non-invasive procedure. The ICA is used to adjust the fuzzy membership functions. The proposed method has been evaluated with the Cleveland and Hungarian datasets. The advantage of this method, compared with others, is the interpretability. The accuracy of the proposed method is 94.92% by 11 rules, and the average length of 4. To compare the colonial competitive algorithm with other metaheuristic algorithms, the proposed method has been implemented with the particle swarm optimization (PSO) algorithm. The results indicate that the colonial competition algorithm is more efficient than the PSO algorithm.