Maximal Itemsets Research Articles

A New Algorithm for Extracting Textual Maximal Frequent Itemsets from Arabic Documents

In this paper, a new technique has been suggested for extracting textual maximal frequent itemsets named Maximal Itemset Miner Algorithm (MIMA). This algorithm begins search process through generating the best initial border in search space depending on minimum support of items in the first level that achieves the general minimum support determined by the user. Our approach for counting itemsets support combines the idea of vertical representation of the data with a queue data structure to store the itemsets. To reduce search space, the algorithm adopted several pruning conditions for each itemsets in the initial border. Experiments performed on standard textual CNN Arabic dataset and proposed method registers less execution time comparing with the Apriori algorithm when applying it on three different size datasets.

Classification of Arabic Documents depending on Maximal Frequent Itemsets

In this paper we introduced techniques for classifying Arabic documents depending on association rules built from maximal frequent itemsets. Parallel Maximal Itemset Miner Algorithm (PMIMA) adopted several conditions to prune search space parallelly introduced for extracting maximal frequent itemsets. Rule length, rule weight and rule majority are three classification methods exploited to classification Arabic documents. Comparing with classification results obtained depending on all frequent itemsets extracted by Apriori, we proved efficiency of ours approach.

Deriving Frequent Itemsets from Lossless Condensed Representation

In data mining, major research topic is frequent itemset mining (FIM). Frequent Itemsets (FIs) usually generating a large amount of Itemsets from database it causing from high memory and long execution time usage. Frequent Closed Itemsets(FCI) and Frequent Maximal Itemsets(FMI) are a reduced lossless representation of frequent itemsets. The FCI allows to decreasing the memory usage and execution time while comparing to FMIs. The whole data of frequent Itemsets(FIs) may be derived from FCIs and FMIs with correct methods. While various study has presented several efficient approach for FCIs and FMIs mining. In sight of this, that we proposed an algorithm called DCFI-Mine for capably derive FIs from Closed FIs and RFMI algorithm derive FMIs to FIs. The advantages of DCFI-Mine algorithm has two features: First, efficiency, different existing algorithm that tends to develop an enormous quantity of Itemsets all through process, DCFI-Mine process the Itemsets straight without candidate generation. But in proposed RFMI multiple scan occurs due to search of item support so efficiency is less than proposed algorithm DCFI-Mine. Second, in terms of losslessness DCFI-Mine and RFMI can discover complete frequent itemset without lapse. Experimental result shows That DCFI-Mine is best deriving FIs in term of memory usage and executions time.

PENERAPAN ALGORITMA MAX-MINER UNTUK ANALISIS POLA BELANJA KONSUMEN (STUDI KASUS KAFELOAJA)

This application is used to produce combination meal for Kafeloaja in order to give the owner advices to make decision in business. In this application the method used are Max-Miner and Market Basket Analysis. Max-Miner is used to find frequent maximal itemset and the result is used to formulate association rule in Market Basket Analysis. The result of this program are menu combinations that can be used by owner to be added as new combination meal.

Reference itemsets: useful itemsets to approximate the representation of frequent itemsets

Deriving frequent itemsets from databases is an important research issue in data mining. The number of frequent itemsets may be unusually large when a low minimum support threshold is given. As such, the design of a compact representation to compress and describe them is an interesting topic. In the past, most related research on compact representation focused on frequent closed itemsets and frequent maximal itemsets. The former is a lossless compact technology that can totally recover all frequent itemsets and their frequencies. Contrarily, the latter may lose some information regarding frequent itemsets, because it reserves frequent itemsets only and is unable to identify their frequency. In this paper, we propose a new compact representation that lies between closed itemsets and maximal itemsets. It can reserve all frequent itemsets and identify their approximate frequency. In addition, an efficient algorithm that corresponds to this new concept is designed to find related key information in databases. Finally, a series of experiments are conducted to show the effectiveness of compact representation and the performance of the proposed algorithm.

The lattice‐based approaches for mining association rules: a review

The traditional methods for mining association rules (ARs) include two phrases: mining frequent itemsets (FIs)/frequent closed itemsets (FCIs)/frequent maximal itemsets (FMIs) and generating ARs from FIs/FCIs/FMIs. Lattice‐based approaches (LBAs) for mining ARs are new approaches including two phrases: frequent itemset lattice (FIL)/frequent closed itemset lattice (FCIL) building and generating ARs from the lattice. Total mining time of LBAs for mining ARs outperforms the traditional methods for mining ARs. Besides, the most important advantage of LBAs for mining ARs is that the algorithms only build the lattice once and mine ARs with many different confidences or many different minimum supports (the thresholds have to be greater than or equal to the threshold used to build lattices) without mining FIs/FCIs again. In this article, we describe a number of existing LBAs for mining ARs on static databases including lattice building and rule generation. In addition, in today's online system, the data often change in several operations such as insert, delete, and update. Hence, a number of LBAs for mining ARs on dynamic databases are mentioned. Finally, complexity analysis of the LBAs for mining ARs is also thoroughly discussed. WIREs Data Mining Knowl Discov 2016, 6:140–151. doi: 10.1002/widm.1181This article is categorized under: Algorithmic Development > Association Rules

A distributed maximal frequent itemset mining with multi agents system on bitmap join indexes selection

The amount of information in a data warehouse tends to be extremely large and queries may involve several complex join and aggregate operations at the same time. By using the right indices, the database administrator can speed up these OLAP queries and dramatically shorten processing times. However, selection of an optimal set of indices is a very hard task because of the exponential number of attribute candidates that can be used in the selection process. Addressing this problem, we propose a new approach with two main phases. The first involves pruning the search space to reduce the number of indices candidates. To that end, we use a distributed maximal itemsets mining approach based on a multi agent system that can significantly reduce the complexity of the selection process. We also incorporate a convertible anti-monotone constraint that contains information on the profit of index. The second phase uses also a multi agent's architecture to select final indices using a subset of attribute candidates. This final configuration will provide benefit to OLAP queries, but will also respect the disk space constraint. We validate our proposed approach using an experimental evaluation.

Speeding up maximal fully-correlated itemsets search in large databases

Finding the most interesting correlations among items is essential for problems in many commercial, medical, and scientific domains. Our previous work on the maximal fully-correlated itemset (MFCI) framework can rule out the itemsets with irrelevant items and its downward-closed property helps to achieve good computational performance. However, to calculate the desired MFCIs in large databases, there are still two computational issues. First, unlike finding maximal frequent itemsets which can start the pruning from 1-itemsets, finding MFCIs must start the pruning from 2-itemsets. When the number of items in a given dataset is large and the support of all the pairs cannot be loaded into the memory, the IO cost (\(O(n^2)\)) for calculating correlation of all the pairs can be very high. Second, users usually need to try different correlation thresholds for different desirable MFCIs. Therefore, the cost of processing the Apriori procedure each time for a different correlation threshold is also very high. Consequently, we proposed two techniques to solve these problems. First, we identify the correlation upper bound for any good correlation measure to avoid unnecessary IO query for the support of pairs, and make use of their common monotone property to prune many pairs even without computing their correlation upper bounds. In addition, we build an enumeration tree to save the fully-correlated value for all the MFCIs under a given initial correlation threshold. We can either efficiently retrieve the desired MFCIs for any given threshold above the initial threshold or incrementally grow the tree if the given threshold is below the initial threshold. Experimental results show that our algorithm can be an order of magnitude faster than the original MFCI algorithm.

BAHUI

Mining high utility itemsets is one of the most important research issues in data mining owing to its ability to consider nonbinary frequency values of items in transactions and different profit values for each item. Although a number of relevant approaches have been proposed in recent years, they incur the problem of producing a large number of candidate itemsets for high utility itemsets. In this paper, the authors propose an efficient algorithm, namely BAHUI (Bitmap-based Algorithm for High Utility Itemsets), for mining high utility itemsets with bitmap database representation. In BAHUI, bitmap is used vertically and horizontally. On the one hand, BAHUI exploits a divide-and-conquer approach to visit itemset lattice by using bitmap vertically. On the other hand, BAHUI horizontally uses bitmap to calculate the real utilities of candidates. Using bitmap compression scheme, BAHUI reduces the memory usage and makes use of the efficient bitwise operation. Furthermore, BAHUI only records candidate high utility itemsets with maximal length, and inherits the pruning and searching strategies from maximal itemset mining problem. Extensive experimental results show that the BAHUI algorithm is both efficient and scalable.

Evaluation and optimization of frequent, closed and maximal association rule based classification

Real world applications of association rule mining have well-known problems of discovering a large number of rules, many of which are not interesting or useful for the application at hand. The algorithms for closed and maximal itemsets mining significantly reduce the volume of rules discovered and complexity associated with the task, but the implications of their use and important differences with respect to the generalization power, precision and recall when used in the classification problem have not been examined. In this paper, we present a systematic evaluation of the association rules discovered from frequent, closed and maximal itemset mining algorithms, combining common data mining and statistical interestingness measures, and outline an appropriate sequence of usage. The experiments are performed using a number of real-world datasets that represent diverse characteristics of data/items, and detailed evaluation of rule sets is provided as a whole and w.r.t individual classes. Empirical results confirm that with a proper combination of data mining and statistical analysis, a large number of non-significant, redundant and contradictive rules can be eliminated while preserving relatively high precision and recall. More importantly, the results reveal the important characteristics and differences between using frequent, closed and maximal itemsets for the classification task, and the effect of incorporating statistical/heuristic measures for optimizing such rule sets. With closed itemset mining already being a preferred choice for complexity and redundancy reduction during rule generation, this study has further confirmed that overall closed itemset based association rules are also of better quality in terms of classification precision and recall, and precision and recall on individual class examples. On the other hand maximal itemset based association rules, that are a subset of closed itemset based rules, show to be insufficient in this regard, and typically will have worse recall and generalization power. Empirical results also show the downfall of using the confidence measure at the start to generate association rules, as typically done within the association rule framework. Removing rules that occur below a certain confidence threshold, will also remove the knowledge of existence of any contradictions in the data to the relatively higher confidence rules, and thus precision can be increased by disregarding contradictive rules prior to application of confidence constraint.

Frequent item set mining

AbstractFrequent item set mining is one of the best known and most popular data mining methods. Originally developed for market basket analysis, it is used nowadays for almost any task that requires discovering regularities between (nominal) variables. This paper provides an overview of the foundations of frequent item set mining, starting from a definition of the basic notions and the core task. It continues by discussing how the search space is structured to avoid redundant search, how it is pruned with the a priori property, and how the output is reduced by confining it to closed or maximal item sets or generators. In addition, it reviews some of the most important algorithmic techniques and data structures that were developed to make the search for frequent item sets as efficient as possible. © 2012 Wiley Periodicals, Inc.This article is categorized under: Algorithmic Development > Association Rules Application Areas > Data Mining Software Tools Technologies > Association Rules

High utility pattern mining using the maximal itemset property and lexicographic tree structures

The problem of high utility mining is discovering all of the high utility itemsets in a transactional database. Most algorithms find high utility itemsets in two steps. The first step identifies all of the potential itemsets. The second step then determines the high utility itemsets from the set of potential itemsets. The large number of potential itemsets in the first step is generally the mining bottleneck. If we can reduce the number of potential itemsets, the mining performance can be improved significantly. In this paper, we use a maximal itemset property and propose an algorithm called UMMI (high Utility Mining using the Maximal Itemset property) to significantly reduce the number of potential itemsets in the first step. In the second step, UMMI uses an effective lexicographic tree structure to determine all of the high utility itemsets. In general, UMMI outperforms all three of the previously used algorithms, including CTU-PRO, an optimized TWU-mining algorithm, and Two-Phase, in our experiments using synthetic datasets. On average, UMMI is 5, 3, and 7 times faster than CTU-PRO, TWU-mining, and Two-Phase, respectively. In a real data experiment, UMMI is 6 times faster than Two-Phase. The other two algorithms are not capable of completing the mining step in a reasonable amount of time. UMMI uses an approximately fixed amount of memory, which is generally less than the other algorithms for each mining. The experimental results show that the proposed algorithm can mine the high utility itemsets efficiently. In addition, UMMI is linearly scalable with respect to the number of transactions.

Efficient Mining Algorithms of Finding Frequent Datasets

This work proposes an efficient mining algorithm to find maximal frequent item sets from relational database. It adapts to large datasets.Itemset is stored in list with special structure. The two main lists called itemset list and Frequent itemset list are created by scanning database once for dividing maximal itemsets into two categories depending on whether the itemsets to achieve minimum support number. Sub itemsets whose superset is in itemset list are generated by recursion to make sure that each sub itemsets appeared before its superset. As current sub itemsets being joined to frequent itemset list, its sub itemsets are pruned from the itemset list. At last, all sub itemsets whose nearest superset is in frequent itemset list are pruned from the frequent itemset list to hold all maximal frequent itemsets.We compare our algorithms and FP-Growth by two sets of time-consuming experiments to prove the superiority of our efficient algorithm both not only with increasing datasets but also with changing mini-support.

Maximal Frequent Itemset Generation Using Segmentation Apporach

Finding frequent itemsets in a data source is a fundamental operation behind Association Rule Mining. Generally, many algorithms use either the bottom-up or top-down approaches for finding these frequent itemsets. When the length of frequent itemsets to be found is large, the traditional algorithms find all the frequent itemsets from 1-length to n-length, which is a difficult process. This problem can be solved by mining only the Maximal Frequent Itemsets (MFS). Maximal Frequent Itemsets are frequent itemsets which have no proper frequent superset. Thus, the generation of only maximal frequent itemsets reduces the number of itemsets and also time needed for the generation of all frequent itemsets as each maximal itemset of length m implies the presence of 2m-2 frequent itemsets. Furthermore, mining only maximal frequent itemset is sufficient in many data mining applications like minimal key discovery and theory extraction. In this paper, we suggest a novel method for finding the maximal frequent itemset from huge data sources using the concept of segmentation of data source and prioritization of segments. Empirical evaluation shows that this method outperforms various other known methods.

Generalized association rule mining using an efficient data structure

The goal of this paper is to use an efficient data structure to find the generalized association rules between the items at different levels in a taxonomy tree under the assumption that the original frequent itemsets and association rules were generated in advance. The primary challenge of designing an efficient mining algorithm is how to make use of the original frequent itemsets and association rules to directly generate new generalized association rules, rather than rescanning the database. In the paper, we used an efficient data structure called the frequent closed enumeration table (FCET) to store the relevant information. It stores only maximal itemsets, and can be used to derive the information of the subset itemsets in a maximal itemset through a hash function. In the proposed algorithms GMAR and GMFI, we used join methods and/or pruning techniques to generate new generalized association rules. Through several comprehensive experiments, we found that both algorithms are much better than BASIC and Cumulate algorithms also using the efficient data structure (FCET), owing to fewer candidate itemsets generated by GMAR and GMFI. Furthermore, the GMAR algorithm prunes a large amount of irrelevant rules based on the minimum confidence.

Meta itemset: a new concise representation of frequent itemset

The sheer size of all frequent itemsets is one challenging problem in data mining research. Based on both closed itemset and maximal itemset, meta itemset which is a new concise representation of frequent itemset is proposed. It is proved that both closed itemset and maximal itemset are special cases of meta itemset. The set of all closed itemsets and the set of all maximal itemsets form the upper bound and the lower bound of the set of all meta itemsets. Then, property and pruning strategies of meta itemset are discussed. Finally, an efficient algorithm for mining meta itemset is proposed. Experimental results show that the proposed algorithm is effective and efficient.

Distributed Mining of Maximal Frequent Itemsets on a Data Grid System

In this paper, we propose a new algorithm, named Grid-based Distributed Max-Miner (GridDMM), for mining maximal frequent itemsets from databases on a Data Grid. A frequent itemset is maximal if none of its supersets is frequent. GridDMM is specifically suitable for use in Grid environments due to low communication and synchronization overhead. GridDMM consists of a local mining phase and a global mining phase. During the local mining phase, each node mines the local database to discover the local maximal frequent itemsets, then they form a set of maximal candidate itemsets for the top-down search in the subsequent global mining phase. A new prefix-tree data structure is developed to facilitate the storage and counting of the global candidate itemsets of different sizes. We built a Data Grid system on a cluster of workstations using the open-source Globus Toolkit, and evaluated the GridDMM algorithm in terms of performance, scalability, and the overhead of communication and synchronization. GridDMM demonstrates better performance than other sequential and parallel algorithms, and its performance is scalable in terms of the database size and the number of nodes.

Generalized domination in closure systems

In the context of extracting maximal item sets and association rules from a binary data base, the graph-theoretic notion of domination was recently used to characterize the neighborhood of a concept in the corresponding lattice. In this paper, we show that the notion of domination can in fact be extended to any closure operator on a finite universe and be efficiently encoded into propositional Horn functions. This generalization enables us to endow notions and algorithms related to Formal Concept Analysis with Horn minimization and minimal covers of functional dependencies in Relational Databases.

An Efficient Graph-Based Method for Parallel Mining Problems

Parallel association rules mining is a noticeable problem in data mining. However, little work has been proposed to deal with three important issues: (1) less memory usage; (2) less communication, among the involved computers, over the network; and (3) load balance among computers. In this paper, we present a graph-based scheme to solve the parallel mining problem by applying independent groups (clusters of maximal cliques). To bring the three issues to a close, the purpose of the independent groups aims at dividing a database into several independent sub-databases, so each sub-database can be employed independently to perform mining algorithms. To emphasis the effectiveness of the graph-based scheme, we adopt the independent groups not only for maximal large itemsets mining but also for general large itemsets mining. The experimental results show that our scheme can improve the efficiency for parallel mining when the independent groups are well-organized and designed.

Design and Implementation of a Generator of Large , Dense ,or Sparse Databases to Test Association Rules Miner

Association rules discovery has emerged as a very important problem in knowledge discovery in database and data mining. A number of algorithms is presented to mine association rules. There are many factors that affect the efficiency of rules mining algorithms, such as largeness, denances, and sparseness of databases used to be mined, in addition to number of items, number and average sizes of transactions, number and average sizes of frequent itemscts, and number and average sizes of potentially maximal itemsets. It is impossible to change present realworld catabase's characteristics to fairly test and determine the best and wurst cases of rule-mining algorithms. to be efficiently used for present and future databases. So the researchers attend to construct artificial database to qualitative and quantitative presence of the above mentioned factors to test the efficiency of rule mining algorithms and programs. The construction of such databases CATmes very large amount of the and efforts. This resent presents a software system, generator, to construct artificial databases.