Research and improvement on Apriori algorithm of association rule mining


 
 
 
 Apriori algorithm is a classical algorithm of association rule mining and widely used for generating frequent item sets. However, the original Apriori algorithm has some limitation such as it needs to scan the dataset many times to discover all frequent itemset and generate huge number of candidate itemset. To overcome these limitations, researchers have made a lot of improvements to the Apriori such as candidate generation, without candidate generation, transaction reduction, partitioning, and sampling. When it comes to mine massive data, these algorithms failed to prove efficiency because limitation of the processing capacity, storage capacity, and main memory constraints. Therefore, parallel and distributed algorithms are developed to perform large-scale computing in ARM on multiple processors. However, the problems with most of the parallel and distributed framework are overheads of managing distributed system, lack of high level parallel programming language, and node failures. Hadoop-MapReduce is an efficient, scalable, and simplified programming model for massive data processing and it also available on cloud environment. Cloud computing offers huge computing resources, and capacities to solve big data challenges. Recently many parallel algorithms have been proposed on Hadoop-MapReduce to enhance the performance of Apriori algorithm but there are some drawbacks: since multiple scan over the dataset is needed to generate candidate itemset, it consume more execution time. The aim of this study is to propose a parallel Transaction Reduction MapReduce Apriori algorithm (TRMR-Apriori) which is reduce unnecessary transaction values and transactions from the dataset in parallel manner to overcome above problems. The experiments show that TRMR-Apriori is able to achieve better execution time to discover frequent itemset those of previous sequential ARM algorithms such as Apriori, AprioriTid, Eclat, and FP-Growth and the previous parallel algorithms such as PApriori, MRApriori, and Modified Apriori with different condition on homogeneous computing environment using Hadoop-MapReduce platform in cloud. Overall, the TRMR-Apriori shows the strength to extract the frequent itemset from massive dataset in cloud.
 
 
 
 

In order to solve these problems how to easily generate candidate frequent item sets and fast compute support of candidate item sets, an algorithm of association rules mining based on binary has been introduced. However, one presented binary mining algorithm is only suitable for mining some relative short frequent item sets since the way of generating candidate item sets is also similar to Apriori, another is only suitable for mining long frequent item sets, which generates candidate item sets by up-down search strategy. And so aiming to mining general frequent item sets, this paper proposes an algorithm of association rules double search mining based on binary, which is different from tradition association rules mining algorithm based on double search strategy. The algorithm doesn’t use combination of set theory to generate candidate item sets but binary logic operation that is also used to compute support of candidate item sets, which can use character digital to reduce the number of scanned transaction. The algorithm gets rid of shortage about some presented algorithms based on binary. The experiment based on above three algorithms indicates that the efficiency about double search strategy is fast and efficient when mining general frequent item sets which aren’t confined.

There are some problems about some optimization algorithms of Apriori such as they consume large memory space although they reduce the numbers of database scanning, or the problem about the difficulties to realize programming. This paper presents an Apriori's optimization algorithm. The algorithm first uses the order character of itemsets to reduce the times of comparison and connection when it connects and generates the candidate itemsets, then compresses the candidate itemsets according to the following situation: whether the number of element "a" in the frequent K-itemsets is less than K. Through the experiment, it is proved that the algorithm can not only realize programming easily but also improve the efficiency of mining association rules.

Discovering frequent item sets is the main way of association rules mining, and it is also the focus of the study in algorithms for association rules mining. The classical Apriori algorithm and its improved algorithms of association rules mining can be generally classified as one based on SQL and the other based on memory. To improve the data-mining efficiency, the authors proposed an efficient algorithm for discovering frequent item sets. After analyzing the efficiency bottlenecks in some algorithms based on memory, the algorithm used a method that could generate and test candidate item sets efficiently to optimize the speed of item sets generation. The experimental results show that the proposed algorithm can assuredly improve the mining efficiency.

In view of the existing defects, based on the analysis of the ideas and methods improved by predecessors, and combining with the characteristics of mobile platform teaching evaluation, the original classic Apriori algorithm is improved. The improved database only needs to scan the original database once, store the transactions belonging to each frequent itemset, and store these collections in the newly added database. Therefore, when counting the candidate item set generated by self-join, only the intersection of the transaction set stored in the subset of the candidate item set is required to obtain the number of supported items in the candidate item set. The improved algorithm can reduce the scanning times of the original database and compress the scanning range, thus improving the efficiency of the algorithm. The teaching evaluation system based on mobile platform meets the needs of all the teachers and students and provides great convenience. Therefore, it can greatly improve the participation rate and make the data more comprehensive and complete.

Owing to the conception of big data and massive data processing there are increasing owes related to the temporal aspects of the data processing. In order to address these issues a continuous progression in data collection, storage technologies, designing and implementing large-scale parallel algorithm for Data mining is seen to be emerging in a rapid pace. In this regards, the Apriori algorithms have a great impact for finding frequent item sets using candidate generation. This paper presents highlights on parallel algorithm for mining association rules using MPI for passing message base in the Master-Slave based structural model.

Under the background of Internet big data era, the amount of data information in all walks of life is increasing day by day. How to filter, mine and manage different kinds of data information has become an important topic of network communication, public service data analysis, privacy protection and other activities. This paper mainly discusses the algorithm of association rules mining based on Apriori algorithm. The association algorithm is used to process the raw data and privacy data of different network data sets, and dig out the multi-level frequent itemsets in the database.

This paper discusses the massive data processing method-Apriori algorithm for mining association rules. The Apriori algorithm repeatedly scans the database for shortcomings. Based on a Hadoop platform to block huge amounts of data, we provide the massive data processing method-Apriori algorithm for mining association rules.

The Apriori algorithm creates all possible association rules between items in the database using the Association Rule Mining and Apriori Algorithm. Using Ant Colony Optimization, a new algorithm is proposed for improving association rule mining results. Using ant colony behaviour as a starting point, an optimization of ant colonies (ACO) is developed. The Apriori algorithm creates association rules. Determine the weakest rule set and reduce the association rules to find rules of higher quality than apriori based on the Ant Colony algorithm's threshold value. Through optimization and improvement of rules generated for ACO, the proposed research work aims to reduce the scanning of datasets.

Discovering the relationship between behavior and personality of learner in the web-based learning environment is a key to guide learners in the learning process. This paper proposes a new concept called personality mining to find the “deep” personality through the observed data about the behavior. First, a learner model which includes personality model and behavior model is proposed. Second, we have designed and implemented an improved algorithm, which is based on Apriori algorithm widely used in market basket analysis, to identify the relationship. Third, we have discussed various issues like constructing the learner model, unifying the value domain of heterogeneous model attributes, and improving Apriori algorithm with decision domain. Experiment result indicated that this algorithm for mining association rules between behavior and personality is feasible and efficient. The algorithm has been used in a web-based learning environment developed at Xi’an Jiaotong University.

Data mining is the process of finding correlations or patterns among dozens of fields in large relational databases. While Association Rules Mining (ARM) algorithm especially the Apriori algorithm has been an active research work in recent years. Diverse improvement varies in term of producing more frequent items and also generating further k-length. The idea is to produce better pattern and more interesting rules. In this paper, we propose new approach for ARM based on Multiple Attribute Value within the non-binary search spaces. The proposed algorithm improves the existing frequent pattern mining by generating the most frequent values (item) within the attribute and generate candidate based on the frequent attribute value. The main idea of our work is to discover more meaningful frequent items and maximum k-length items. The experimental results show that our proposed MAV frequent pattern mining enhance the impact in generating more frequents items and maximum length.

Though the association rule mining algorithms has simple ideas, their computations are large. It may cause that the existing frequent item set increases exponentially with the increasing of items number. Many sampling algorithms have low accuracy and they need to solve a series of NP hard problems. Therefore, based on the research of existing sampling algorithms, a new hierarchical dichotomy sampling algorithm is proposed. It performs sampling on existing data before each mining. The frequent k-item set can be partitioned as averagely as possible during the average partition of data, to ensure the rules with high accuracy can be mined through fewer data. The experiments verify that the improved algorithm has superior performance on accuracy, compared to traditional algorithm, and it also shows advantage in the running time.

Although the FP-Growth association-rule mining algorithm is more efficient than the Apriori algorithm, it has two disadvantages. The first is that the FP-tree can become too large to be created in memory; the second is the serial processing approach used. In this paper, a kind of parallel association-rule mining algorithm has been proposed. It does not need to create an overall FP-tree, and it can distribute data mining tasks over several computing nodes to achieve parallel processing. This approach will greatly improve efficiency and processing ability when used for mining association rules and is suitable for association-rule mining on massive data sets.

Association rules mining is one of the most important problems in knowledge discovery and data mining. The goal of it is to acquire consumption habits of customers by discovering the relationships between items from a transaction database that has a large number of transactions and items. The most compute intensive process for ARM is to find out the frequent 1-itemsets and 2-itemsets. In this paper, we propose a quantum algorithm for finding out the frequent 1-itemsets and 2-itemsets. In our algorithm, to mine the frequent 1-itemsets efficiently, we use the technique of amplitude amplification. To mine the frequent 2-itemsets efficiently, we propose a new tomography scheme, i.e., pure-state-based quantum state tomography. It is shown that our algorithm is potential to achieve exponential speedup in the number of transactions and polynomial speedup in the number of items over the classical algorithm.

Among the many mining algorithms of association rules, Apriori Algorithm is a classical algorithm that has caused the most discussion; it can effectively carry out the mining association rules. However, based on Apriori Algorithm, most of the traditional algorithms existed "item sets generation bottleneck" problem, and are very time-consuming. An enhance algorithm associating which is based on the user interest and the importance of itemsets is put forward by the paper, incorporate item that user is interested in into the itemsets as a seed item, then scan the database, incorporate all other items which are in the same transaction into itemsets, Construct user interest itemsets, reduce unnecessary itemsets; through the design of the support functions algorithm not only considered the frequency of itemsets, but also consider different importance between different itemsets. The new algorithm reduces the storage space, improves the efficiency and accuracy of the algorithm.