Itemsets In Transaction Database Research Articles

HEPM: High-efficiency pattern mining

Pattern mining (PM) is an important field of data mining and has gained considerable momentum recently, mainly owing to the massive growth of big data. PM often sets attentive objectives such as mining frequent or high utility patterns to obtain attractive patterns. High utility patterns address the defect of frequent patterns that cannot reveal the maximum profit. However, it neglects another vital factor, cost or investment. This paper proposes a new high-efficiency PM problem that considers both utility and investment. The problem aims to find patterns with the maximum profit-to-investment ratio. Our paper is devoted to studying high-efficiency itemsets in transaction databases. We first formulate the criteria for a high-efficiency PM problem. Subsequently, we propose a two-phase algorithm called HEPM and an improved one-phase algorithm called HEPMiner to discover high-efficiency patterns in a transaction database. We design a corresponding pruning strategy within HEPM to reduce the search space. In HEPMiner, we utilize a novel efficiency-list and an estimated efficiency co-occurrence structure in the pruning strategies to further improve the mining performance. Moreover, we derive the upper bounds of efficiency for both algorithms. The experimental results demonstrate the effectiveness and efficiency of our two algorithms.

LCM from FCA point of view: A CbO-style algorithm with speed-up features

LCM is an algorithm for enumeration of frequent closed itemsets in transaction databases. It is well known that when we ignore the required frequency, the closed itemsets are exactly intents of formal concepts in Formal Concept Analysis (FCA). We describe LCM in terms of FCA and show that LCM is basically the Close-by-One algorithm with multiple speed-up features for processing sparse data. We analyze the speed-up features and compare them with those of similar FCA algorithms, like FCbO and algorithms from the In-Close family.

A new approximate method for mining frequent itemsets from big data

Mining frequent itemsets in transaction databases is an important task in many applications. It becomes more challenging when dealing with a large transaction database because traditional algorithms are not scalable due to the limited main memory. In this paper, we propose a new approach for the approximately mining of frequent itemsets in a big transaction database. Our approach is suitable for mining big transaction databases since it uses the frequent itemsets from a subset of the entire database to approximate the result of the whole data, and can be implemented in a distributed environment. Our algorithm is able to efficiently produce high-accurate results, however it misses some true frequent itemsets. To address this problem and reduce the number of false negative frequent itemsets we introduce an additional parameter to the algorithm to discover most of the frequent itemsets contained in the entire data set. In this article, we show an empirical evaluation of the results of the proposed approach.

TT-Miner: Topology-Transaction Miner for Mining Closed Itemset

Mining frequent closed itemsets (FCIs) from transaction databases is a fundamental problem in many data mining applications. All the enumeration algorithms enumerate FCIs by adding a singleton item to an itemset and then checking whether it is closure. In order to reduce enumerations, we enumerate FCIs by adding an itemset to the existing FCI itemset. To this end, we first analyze a binary relation on the set of itemsets in transaction databases, show co-occurrence relation among items, and prove that the relation is reflexive and transitive. Next, we use the relation to construct a topology for the set of itemsets and prove that all FCIs are included in the topology. Then, we construct a topology-transaction tree (TT-tree) and provide a topology-transaction miner (TT-Miner) algorithm for enumerating FCIs in the TT-tree. Finally, an extensive experimental evaluation on a number of real and synthetic databases shows that the TT-Miner is an efficient and scalable algorithm compared with the previous methods.

Mining Top-k Regular High-Utility Itemsets in Transactional Databases

Mining high-utility itemsets is an important task in the area of data mining. It involves exponential mining space and returns a very large number of high-utility itemsets. In a real-time scenario, it is often sufficient to mine a small number of high-utility itemsets based on user-specified interestingness. Recently, the temporal regularity of an itemset is considered as an important interesting criterion for many applications. Methods for finding the regular high utility itemsets suffers from setting the threshold value. To address this problem, a novel algorithm called as TKRHU (Top k Regular High Utility Itemset) Miner is proposed to mine top-k high utility itemsets that appears regularly where k represents the desired number of regular high itemsets. A novel list structure RUL and efficient pruning techniques are developed to discover the top-k regular itemsets with high profit. Efficient pruning techniques are designed for reducing search space. Experimental results show that proposed algorithm using novel list structure achieves high efficiency in terms of runtime and space.

Exploiting GPU and cluster parallelism in single scan frequent itemset mining

This paper considers discovering frequent itemsets in transactional databases and addresses the time complexity problem by using high performance computing (HPC). Three HPC versions of the Single Scan (SS) algorithm are proposed. The first one (GSS) implements SS on a GPU (Graphics Processing Unit) architecture using an efficient mapping between thread blocks and the input data. The second approach (CSS) implements SS on a cluster architecture by scheduling independent jobs to workers in a cluster. The third, (CGSS) accelerates the frequent itemset mining process by using multiple cluster nodes equipped with GPUs. Moreover, three partitioning strategies are proposed to reduce GPU thread divergence and cluster load imbalance. Results show that CGSS outperforms SS, GSS, and CSS in terms of speedup. Specifically, CGSS provides up to a 350 times speedup for low minimum support values on large datasets. GCSS demonstrably outperforms the state-of-the-art HPC-based algorithms on big databases.

Efficient high utility itemset mining using buffered utility-lists

Discovering high utility itemsets in transaction databases is a key task for studying the behavior of customers. It consists of finding groups of items bought together that yield a high profit. Several algorithms have been proposed to mine high utility itemsets using various approaches and more or less complex data structures. Among existing algorithms, one-phase algorithms employing the utility-list structure have shown to be the most efficient. In recent years, the simplicity of the utility-list structure has led to the development of numerous utility-list based algorithms for various tasks related to utility mining. However, a major limitation of utility-list based algorithms is that creating and maintaining utility-lists are time consuming and can consume a huge amount of memory. The reasons are that numerous utility lists are built and that the utility-list intersection/join operation to construct a utility-list is costly. This paper addresses this issue by proposing an improved utility-list structure called utility-list buffer to reduce the memory consumption and speed up the join operation. This structure is integrated into a novel algorithm named ULB-Miner (Utility-List Buffer for high utility itemset Miner), which introduces several new ideas to more efficiently discover high utility itemsets. ULB-Miner uses the designed utility-list buffer structure to efficiently store and retrieve utility-lists, and reuse memory during the mining process. Moreover, the paper also introduces a linear time method for constructing utility-list segments in a utility-list buffer. An extensive experimental study on various datasets shows that the proposed algorithm relying on the novel utility-list buffer structure is highly efficient in terms of both execution time and memory consumption. The ULB-Miner algorithm is up to 10 times faster than the FHM and HUI-Miner algorithms and consumes up to 6 times less memory. Moreover, it performs well on both dense and sparse datasets.

Parallel Mining for High Utility Itemsets Mining by Efficient Data Structure

Mining high utility itemsets in transaction database is an important task in data mining and widely applied in many areas. Recently, many algorithms have been proposed, but most algorithms for identifying high utility itemsets need to generate candidate sets by overestimating their utility and then calculating their exact utility value. Therefore, the number of candidate itemsets is much larger than the actual number of high utility itemsets. In this paper, we introduce the Retail Transaction-Weighted Utility (RTWU) structure and propose two algorithms: EAHUIMiner algorithm and PEAHUI-Miner parallel algorithm. They have been experimented and compared to the two most efficient algorithms: EFIM and FHM. Results show that our algorithm is better with sparse datasets. DOI: 10.32913/rd-ict.vol3.no14.519

A two-phase approach to mine short-period high-utility itemsets in transactional databases

The discovery of high-utility itemsets (HUIs) in transactional databases has attracted much interest from researchers in recent years since it can uncover hidden information that is useful for decision making, and it is widely used in many domains. Nonetheless, traditional methods for high-utility itemset mining (HUIM) utilize the utility measure as sole criterion to determine which item/sets should be presented to the user. These methods ignore the timestamps of transactions and do not consider the period constraint. Hence, these algorithms often finds HUIs that are profitable but that seldom occur in transactions. In this paper, we address this limitation of previous methods by pushing the period constraint in the HUI mining process. A new framework called short-period high-utility itemset mining (SPHUIM) is designed to identify patterns in a transactional database that appear regularly, are profitable, and also yield a high utility under the period constraint. The aim of discovering short-period high-utility itemsets (SPHUI) is hence to identify patterns that are interesting both in terms of period and utility. The paper proposes a baseline two-phase short-period high-utility itemset (SPHUITP) mining algorithm to mine SPHUIs in a level-wise manner. Then, to reduce the search space of the SPHUITP algorithm and speed up the discovery of SPHUIs, two pruning strategies are developed and integrated in the baseline algorithm. The resulting algorithms are denoted as SPHUIMT and SPHUITID, respectively. Substantial experiments both on real-life and synthetic datasets show that the three proposed algorithms can efficiently and effectively discover the complete set of SPHUIs, and that considering the short-period constraint and the utility measure can greatly reduce the number of patterns found.

Efficiently mining frequent itemsets in transactional databases

Discovering frequent itemsets is an essential task in association rules mining and it is considered to be computationally expensive. To find the frequent itemsets, the algorithm of frequent pattern growth (FP-growth) is one of the best algorithms for mining frequent patterns. However, many experimental results have shown that building conditional FP-trees during mining data using this FP-growth method will consume most of CPU time. In addition, it requires a lot of space to save the FP-trees. This paper presents a new approach for mining frequent item sets from a transactional database without building the conditional FP-trees. Thus, lots of computing time and memory space can be saved. Experimental results indicate that our method can reduce lots of running time and memory usage based on the datasets obtained from the FIMI repository website.

BitTableFI: An efficient mining frequent itemsets algorithm

Mining frequent itemsets in transaction databases, time-series databases and many other kinds of databases is an important task and has been studied popularly in data mining research. The problem of mining frequent itemsets can be solved by constructing a candidate set of itemsets first, and then, identifying those itemsets that meet the frequent itemset requirement within this candidate set. Most of the previous research mainly focuses on pruning to reduce the candidate itemsets amounts and the times of scanning databases. However, many algorithms adopt an Apriori-like candidate itemsets generation and support count approach that is the most time-wasted process. To address this issue, the paper proposes an effective algorithm named as BitTableFI. In the algorithm, a special data structure BitTable is used horizontally and vertically to compress database for quick candidate itemsets generation and support count, respectively. The algorithm can also be used in many Apriori-like algorithms to improve the performance. Experiments with both synthetic and real databases show that BitTableFI outperforms Apriori and CBAR which uses ClusterTable for quick support count.