• Annual Conference of KIPS
• /
• 2019.05a
• /
• pp.33-35
• /
• 2019

서로 다른 아이템 집합의 연관성을 분석하는 것을 연관규칙분석이라 한다. 대표적인 알고리즘으로 Apriori 알고리즘이 있지만 DB스캔 횟수가 많아질 수 있고 후보 집합 생성으로 인해서 속도가 느려질 수 있다는 단점이 있다. 이를 효율적으로 개선한 FP-Growth 알고리즘을 구현하여 임의의 데이터를 이용하여 알고리즘의 속도에 대해 연구한다.

• Journal of Digital Convergence
• /
• v.15 no.6
• /
• pp.229-236
• /
• 2017

An efficient frequent pattern algorithm is essential for mining association rules as well as many other mining tasks for convergence with its application spread over a very broad spectrum. Models for mining pattern have been proposed using a FP-tree for storing compressed information about frequent patterns. In this paper, we propose a centroid frequent pattern growth algorithm which we called "CAWFP-Growth" that enhances he FP-Growth algorithm by making the center of weights and frequencies for the itemsets. Because the conventional constraint of maximum weighted support is not necessary to maintain the downward closure property, it is more likely to reduce the search time and the information loss of the frequent patterns. The experimental results show that the proposed algorithm achieves better performance than other algorithms without scarifying the accuracy and increasing the processing time via the centroid of the items. The MapReduce framework model is provided to handle large amounts of data via a pseudo-distributed computing environment. In addition, the modeling of the proposed algorithm is required in the fully distributed mode.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10c
• /
• pp.259-261
• /
• 2002

데이터마이닝 분야에서 빈발항목집합 탐사에 관한 연구는 활발히 진행되어 왔지만 여전히 많은 메모리 공간과 시간을 필요로 한다. 특히 apriori 알고리즘에 기반한 방법들은 긴 패턴이 생성될수록 지수적으로 시간과 공간이 증가한다. 최근에 발표된 fp-growth는 일반적인 데이터 집합에서 우수한 성능을 보이나 희소 데이터 집합에서 효율적인 성능을 보여주지 못한다. 본 논문에서는 길이가 2인 빈발항목집합 L2에 기반한 L2-tree 구조를 제안한다. 또한 L2-tree에서 빈발항목집합을 탐사하는 L2-traverse 알고리즘을 제안한다. L2-tree는 L2를 기반으로 하기 때문에 L2가 상대적으로 적은 희소 데이터 집합 환경에서 적은 메모리 공간을 사용하게 된다. L2-traverse 알고리즘은 별도의 추출 데이터베이스를 생성하는 FP-growth와 달리 단순히 L2-tree를 오직 한번의 깊이 우선 탐사를 통해 빈발항목집합을 찾는다. 최적화 기법으로써 길이가 3인 빈발항목집합 L3가 되지 않는 L2 패턴들을 미리 제거하는 방법으로 C3-traverse 알고리즘을 제안하며 실험을 통해 기존 알고리즘과 비교 검증한다.

• 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.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.12
• /
• pp.623-634
• /
• 2016

Most of existing frequent pattern mining methods address time efficiency and greatly rely on the primary memory. However, in the era of big data, the size of real-world databases to mined is exponentially increasing, and hence the primary memory is not sufficient enough to mine for frequent patterns from large real-world data sets. To solve this problem, there are some researches for frequent pattern mining method based on disk, but the processing time compared to the memory based methods took very time consuming. There are some researches to improve scalability of frequent pattern mining, but their processes are very time consuming compare to the memory based methods. In this paper, we present PPFP as a novel disk-based approach for mining frequent itemset from big data; and hence we reduced the main memory size bottleneck. PPFP algorithm is based on FP-growth method which is one of the most popular and efficient frequent pattern mining approaches. The mining with PPFP consists of two setps. (1) Constructing an IFP-tree: After construct FP-tree, we assign index number for each node in FP-tree with novel index numbering method, and then insert the indexed FP-tree (IFP-tree) into disk as IFP-table. (2) Mining frequent patterns with PPFP: Mine frequent patterns by expending patterns using stack based PUSH-POP method (PPFP method). Through this new approach, by using a very small amount of memory for recursive and time consuming operation in mining process, we improved the scalability and time efficiency of the frequent pattern mining. And the reported test results demonstrate them.

• Smart Media Journal
• /
• v.5 no.4
• /
• pp.75-82
• /
• 2016

Although Korea's economic and domestic automobile market through the change of road environment are growth, the traffic accident rate has also increased, and the casualties is at a serious level. For this reason, the government is establishing and promoting policies to open traffic accident data and solve problems. In this paper, describe the method of predicting traffic accidents by eliminating the class imbalance using the traffic accident data and constructing the Hybrid Model. Using the original traffic accident data and the sampled data as learning data which use FP-Growth algorithm it learn patterns associated with traffic accident injury severity. Accordingly, In this paper purpose a method for predicting the severity of a victim of a traffic accident by analyzing the association patterns of two learning data, we can extract the same related patterns, when a decision tree and multinomial logistic regression analysis are performed, a hybrid model is constructed by assigning weights to related attributes.

• Journal of Korea Spatial Information System Society
• /
• v.12 no.1
• /
• pp.49-56
• /
• 2010

Many researches of frequent pattern mining technique for detecting unknown patterns on spatial data have studied actively. Existing data structures have classified into tree-structure and array-structure, and those structures show the weakness of performance on dense or sparse data. Since spatial data have obtained the characteristics of dense and sparse patterns, it is important for us to mine quickly dense and sparse patterns using only single algorithm. In this paper, we propose novel data structure as compressed patricia frequent pattern tree and frequent pattern mining algorithm based on proposed data structure which can detect frequent patterns quickly in terms of both dense and sparse frequent patterns mining. In our experimental result, proposed algorithm proves about 10 times faster than existing FP-Growth algorithm on both dense and sparse data.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2004.05a
• /
• pp.636-639
• /
• 2004

본 논문은 웹 도큐먼트를 기반으로 사용자에게 의미 있는 정보를 찾아주기 위한 연관 객체 추출 기법인 PMPL(Personalized Multi-Strategey Pattern Loaming) 시스템을 제안하고자 한다. PMPL 모듈은 인터넷의 정보를 여과하여 필터링하고, 사용자 개인화의 키워드를 중심으로 연관된 객체를 추출한다. 이때 연관된 객체 추출 시 대용량 데이터에서 시간적, 공간적면에서 효율적인 연관 탐색 기법인 Fp-Tree와 Fp-Growth 알고리즘을 적용시켰으며, 연관규칙 탐색을 보완하기 위해 가중치 기법인 만유인력 기법을 적용시켰다. PMPL 시스템을 실행한 결과 개인화된 사용자 중심어 기초로 기존의 단일 학습 기법에 비해 더 많은 의미 있는 연관 지식을 추출한 결과가 보였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.337-339
• /
• 2002

연관규칙 탐사기법은 거래(사건) 속에 포함된 품목(항목)간의 연관관계를 발견하고자 할 때 사용하는 기법이며, 독특한 형태의 자료구조를 사용하는 다양한 연관규칙 알고리즘들이 제안되었다. 다양한 특성을 갖는 대용량의 데이터에 대해 효율적으로 연관규칙 탐사를 수행하기 위해서는 저장공간과 실행시간을 모두 고려해야 한다. 본 논문에서는 후보항목집합 발생과정 없이 압축빈발항목집합과 동적링크집합을 이용하여 저장공간 축소와 실행시간 개선을 동시에 고려한 연관규칙 알고리즘을 제안하며, 그 우수성을 증명하기 위해 연관규칙 탐사의 대표적인 자료 구조인 FP-struct, H-Struct와의 저장공간 비교 및 이들 저장구조를 사용하는 FP-growth, H-mine 알고리즘과의 실행시간을 비교한다.

• 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.