• Journal of Korea Multimedia Society
• /
• v.14 no.10
• /
• pp.1323-1334
• /
• 2011

This paper suggests the border based HSFI algorithm to hide sensitive frequent itemsets. Node formation of FP-Tree which is different from the previous one uses the border to minimize the impacts of nonsensitive frequent itemsets in hiding process, including the organization of sensitive and border information, and all transaction as well. As a result of applying HSFI algorithms, it is possible to be the example transaction database, by significantly reducing the lost items, it turns out that HSFI algorithm is more effective than the existing algorithm for maintaining the quality of more improved database.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04c
• /
• pp.392-394
• /
• 2003

연관 규칙 마이닝은 일반적으로 않은 빈발항목집합과 연관 규칙을 생성하며, 생성된 연관 규칙은 상호 포함관계에 있거나 중복되는 경우가 많다. 이는 효과적인 마이닝 뿐 아니라 마이닝의 활용 효용성을 떨어뜨린다. 이를 해결하기 위하여 연관 규칙 마이닝과 동일한 성능을 가지며 생성되는 규칙의 수를 줄일 수 있는 빈발 폐쇄 항목집합 마이닝이 제안되었다. 본 연구에서는 연관규칙 마이닝 방법 중 가장 우수한 성능을 가지는 ARCS 알고리즘을 개선한 빈발 폐쇄 항목집단 마이닝을 제안한다.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.4
• /
• pp.179-189
• /
• 2007

Every purchase a customer makes builds patterns about how products are purchased together. The process of finding these patterns, called market basket analysis, is composed of two steps in the Microsoft Association Algorithm. The first step is to find frequent item-sets. The second step which requires much less time than the first step does is to generate association rules based on frequent item-sets. Even though the first step, finding frequent item-sets, is the core part of market basket analysis, when applied to Online Analytical Processing(OLAP) cubes it always raises several points such as longitudinal analysis becomes impossible and many unpractical transactions are built up. In this paper, a new OLAP cube structures designing method which makes longitudinal analysis be possible and also makes only real customers' purchase patterns be identified is proposed for market basket analysis.

• The KIPS Transactions:PartD
• /
• v.18D no.3
• /
• pp.169-178
• /
• 2011

Recently, data sharing between enterprises or organizations is required matter for task cooperation. In this process, when the enterprise opens its database to the affiliates, it can be occurred to problem leaked sensitive information. To resolve this problem it is needed to hide sensitive information from the database. Previous research hiding sensitive information applied different heuristic algorithms to maintain quality of the database. But there have been few studies analyzing the effects on the items modified during the hiding process and trying to minimize the hided items. This paper suggests eFP-Tree(Extended Frequent Pattern Tree) based FP-Tree(Frequent Pattern Tree) to hide sensitive frequent itemsets. Node formation of eFP-Tree uses border to minimize impacts of non sensitive frequent itemsets in hiding process, by organizing all transaction, sensitive and border information differently to before. As a result to apply eFP-Tree to the example transaction database, the lost items were less than 10%, proving it is more effective than the existing algorithm and maintain the quality of database to the optimal.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.04b
• /
• pp.136-139
• /
• 2002

연관규칙 마이닝과정에서의 빈발항목 탐색의 대표적인 방법으로 알려진 Apriori 알고리즘의 성능을 향상시키기 위한 많은 연구가 진행되어 왔다. 본 논문에서는 트랜잭션 데이터베이스(TDB)에서 생성되는 각 패스의 k-itemset들에 대해 각각 트랜잭션 ID List(TIDist)를 유지하고 이를 이용해 (k+1)-itemset을 효율적으로 찾아내는 방법을 제안한다. 이 방법은 frequent (k+1)-itemset(k>0)의 빈도수 및 TIDList를 TDB 에 대한 스캔이 전혀 없이 k-itemset의 TIDList로부터 직접 구한다. 이는 빈발항목집합을 찾기 위한 탐색 complexity는 크게 줄여줄 뿐 아니라 시간 변화에 따른 빈발항목집합의 분포 정보를 제공해 준다.

• The Journal of the Korea Contents Association
• /
• v.11 no.1
• /
• pp.56-64
• /
• 2011

It is difficult to discover frequent itemsets based on whole data from data stream since data stream has the characteristics of infinity and continuity. Therefore, a specialized data mining method, which reflects the properties of data and the requirement of users, is required. In this paper, we propose the method of FIMWB discovering the frequent itemsets which are reflecting the property that the recent events are more important than old events. Data stream is splitted into batches according to the given time interval. Our method gives a weighted value to each batch. It reflects user's interestedness for recent events. FP-Digraph discovers the frequent itemsets by using the result of FIMWB. Experimental result shows that FIMWB can reduce the generation of useless items and FP-Digraph method shows that it is suitable for real-time environment in comparison to a method based on a tree(FP-Tree).

• The KIPS Transactions:PartD
• /
• v.15D no.4
• /
• pp.475-484
• /
• 2008

Nowadays, lots of data mining applications based on continuous and online real time are increasing by the rapid growth of the data processing technique. In order to do association rule mining in that application, we have to use new techniques to find the frequent itemsets. Most of the existing techniques to find the frequent itemsets should scan the total database repeatedly. But in the application based on the continuous and online real time, it is impossible to scan the total database repeatedly. We have to find the frequent itemsets with only one scan of the data interval for that kind of application. So in this paper we propose an approximation technique which finds the maximum size of the frequent itemsets and items included in the maximum size of the frequent itemsets for the processing of association rule mining.

• The KIPS Transactions:PartD
• /
• v.15D no.6
• /
• pp.741-754
• /
• 2008

Due to the characteristics of a data stream, it is very important to confine the memory usage of a data mining process regardless of the amount of information generated in the data stream. For this purpose, this paper proposes the Prime pattern tree(PPT) for finding frequent itemsets over data streams with using the confined memory space. Unlike a prefix tree, a node of a PPT can maintain the information necessary to estimate the current supports of several itemsets together. The length of items in a prime pattern can be reduced the total number of nodes and controlled by split_delta $S_{\delta}$. The size and the accuracy of the PPT is determined by $S_{\delta}$. The accuracy is better as the value of $S_{\delta}$ is smaller since the value of $S_{\delta}$ is large, many itemsets are estimated their frequencies. So it is important to consider trade-off between the size of a PPT and the accuracy of the mining result. Based on this characteristic, the size and the accuracy of the PPT can be flexibly controlled by merging or splitting nodes in a mining process. For finding all frequent itemsets over the data stream, this paper proposes a PPT to replace the role of a prefix tree in the estDec method which was proposed as a previous work. It is efficient to optimize the memory usage for finding frequent itemsets over a data stream in confined memory space. Finally, the performance of the proposed method is analyzed by a series of experiments to identify its various characteristics.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.05b
• /
• pp.126-129
• /
• 2003

최근 대용량의 데이터베이스로부터 연관규칙을 발견하여 이를 활용하는 단계에서 이러한 연관규칙을 수량항목에도 적용할 수 있도록 확장하는 연구가 소개되고 있다. 본 논문에서는 수량 항목을 이진항목으로 변환하기 위하여 빈발구간 항목집합(Large Interval Itemsets)을 생성할 때 수량 항목이 특정 영역에 집중하여 발생하거나 골고루 분포되어 있지 않은 경우, 이러한 지역성(locality)을 고려하여 빈발구간 항목집합을 생성하는 방법을 제안한다. 이 방법은 기존의 방법보다 많은 수의 세밀한 빈발구간 항목들을 생성할 수 있을 뿐만 아니라 의미 있는 구간을 중심으로 빈발구간 항목들이 순서대로 생성되기 때문에 세밀도를 판단하여 활용할 수 있으며, 원 데이터가 가지고 있는 특성의 손실을 최소화할 수 있는 특징이 있다 또한 인구센서스등 실 데이터를 사용한 성능평가를 통하여 기존의 방법보다 우수함을 보였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04a
• /
• pp.695-697
• /
• 2003

다중데이터베이스 마이닝에서 하나의 데이터 집합을 형성하는 작업은 많은 부하가 따른다. 그러므로, 본 논문에서는, 가중치 거리를 이용한 클러스터링을 통해 관련성이 높은 데이터베이스를 식별하는 기법을 제안한다. 제안한 기법은 빈발한 항목으로 구성된 데이터 집합을 생성하여 데이터베이스 사이의 유사성과 거리를 측정하고 데이터베이스간의 거리에 대한 식별성을 향상시키기 위하여 최다 빈발항목에 대한 비교 연산을 통해 가중치를 산출한다. 그리고 성능평가를 통하여 제안한 기법이 Ideal&Goodness 기법보다 다중데이터베이스의 트랜잭션 데이터베이스에 대한 식별 능력이 우수함을 알 수 있었다.