Discovery Of Association Rules Research Articles

A novel artificial immune system-based approach for mining associative classification rules with stock trading data

Stock market prediction with high accuracy has always been an interesting subject for most investors and professional analysts. Data mining techniques are providing great aid to extract interesting and hidden knowledge from datasets. Financial data mining tools assist investors in their investment decisions, thereby reducing their investment risks. Associative classification rule mining is a promising approach in data mining that utilises the association rule discovery techniques to construct classification systems, also known as associative classifiers. This paper aims to develop an intelligent transaction system based on associative classification rule mining (ACR) and phenotypic artificial immune system (AIS) which discovers trading rules from numerical indicators. A new fitness function as a different measure of quality for quantitative association is suggested considering interestingness of rules. Based on the empirical studies on the top eight companies in the S&P 500 stocks, observed results demonstrate the superior prediction accuracy over the genetic algorithm based technique and the 'buy and hold' strategy.

Efficient attribute selection strategies for association rule mining in high dimensional data

This paper presents a new computational approach to discover interesting relations between variables, called association rules, in large and high dimensional datasets. State-of-the-art techniques are computationally expensive due to reasons like high dimensions, generation of huge number of candidate sets and multiple database scans. In general, most of the enormous discovered patterns are obvious, redundant or uninteresting to the user. So the context of this paper is to improve apriori algorithm to find association rules pertaining to only important attributes from high dimensional data. We employ an information theoretic method together with the concept of QR decomposition to represent the data in its proper substructure form without losing its semantics, by identifying significant attributes. Experiment on real datasets and comparison with the existing technique reveals that the proposed strategy is computationally always faster and statistically always comparable with the apriori algorithms in terms of rules generated and time complexity.

Analysis of Human Mobility Based on Cellular Data

Nowadays not only adult but even teenager and children have then own mobile phones. This phenomena indicates that the mobile phone becomes an important part of everyday’s life. Based on these indication, the amount of cellular data also increased rapidly. Cellular data defined as the data that records communication among mobile phone users. Cellular data is easy to obtain because the telecommunications company had made a record of the data for the billing system of the company. Billing data keeps a log of the users cellular data usage each time. We can obtained information from the data about communication between users. Through data visualization process, an interesting pattern can be seen in the raw cellular data, so that users can obtain prior knowledge to perform data analysis. Cellular data processing can be done using data mining to find out human mobility patterns and on the existing data. In this paper, we use frequent pattern mining and finding association rules to observe the relation between attributes in cellular data and then visualize them. We used weka tools for finding the rules in stage of data mining. Generally, the utilization of cellular data can provide supporting information for the decision making process and become a data support to provide solutions and information needed by the decision makers.

New Algorithms for Data Mining on Grid Computing

Data Mining on the grid is a new challenge which consists of processing a large quantity of data by knowledge discovery methods. New methods which are able to take into consideration all the specificities of data Mining problems as well as those of grid computing (data grid) are needed for a better exploitation of the treatment of available resources in grid computing. In this paper, we have proposed a method for the extraction of association rules, adapted for parallel and distributed environment such as grids. The goal here is to minimize the execution time by reducing the cost of parallelization and communication. A parallel version of the Partition algorithm for finding association rules has been proposed. The introduction of an intelligent distribution of the base on the grid using clustering to minimize the search space in each machine and consequently reduce the processing time has been considered. To this end, another parallel clustering algorithm for treating numerical and categorical data in a distributed environment has been proposed.

Incremental topological spatial association rule mining and clustering from geographical datasets using probabilistic approach

Due to the dynamic updating of real time spatial databases, the preservation of spatial association rules for dynamic database is a vital issue because the updates may not only invalidate some existing rules but also make other rules relevant. Consequently, the dynamic updating of spatial rules was handled by many researchers through the incremental association rule mining algorithm. Accordingly, in this paper we have developed an incremental topological association rule mining of geographical datasets using probabilistic approach. Initially, the spatial database is read out and it is passed through probability-based incremental association rule discovery algorithm to mine the topological spatial association rules. Once the rules are mined from the spatial database, the assumption here is that the database is dynamically updating for every time interval. In order to handle this dynamic nature, the proposed incremental topological association rule mining process is used in this paper. Here, the candidate topological rule generation is done from the spatial association rules using the topological relations such as, nearby, disjoint, intersects and inside/outside and the topological support is calculated using the proposed probabilistic topological support model. Finally, the spatial clustering is performed based on the mined spatial rules. From the experimentation, we proved that the maximum accuracy reached by the proposed method is 83.14% which is higher than the existing methods, which is defined as the ratio of the occurred rules and total number of topological data objects.

An Improved Design for Association Rule Discovery System for Decision Support System

An Improved Design for Association Rule Discovery System for Decision Support System

An efficient dynamic superset bit-vector approach for mining frequent closed itemsets and their lattice structure

Fast discovery of association rules from millions of transactions in a variety of large databases has now become a major challenge in data mining domain. Frequent itemsets and frequent closed itemsets are key sources of mining association rules. Association rules can be mined efficiently from lattice of itemsets. Non-redundancy, accuracy, time efficiency and memory usage are the factors those need to be considered while developing algorithms in order to extract meaningful association rules. In this paper, we propose an efficient bit-vector approach that exploits dual properties, tidset and superset information of an itemset in order to mine frequent closed itemsets with their lattice structure. We introduce a new memory efficient data structure called dynamic superset bit-vector to establish the relationship among frequent closed itemsets in a lattice. The novelty of our approach is that it effectively uses dual data structures called a dynamic bit-vector and a dynamic superset bit-vector jointly in order to reduce the search space and eliminate the generators of non-closed itemsets. Our proposed approach efficiently builds up the subset-superset relationship among the closed itemsets of the lattice structure in a bottom-up manner. Extensive experiments using real-life datasets and pervasive performance comparison with the existing works prove the efficiency and scalability of our proposed approach.

Comparative Evaluation of the Different Data Mining Techniques Used for the Medical Database

Abstract Data mining is the upcoming research area to solve various problems. Classification and finding association are two main steps in the field of data mining. In this paper, we use three classification algorithms: J48 (an open source Java implementation of C4.5 algorithm), Multilayer Perceptron - MLP (a modification of the standard linear perceptron) and Naïve Bayes (based on Bayes rule and a set of conditional independence assumptions) of the Weka interface. These classifiers have been used to choose the best algorithm based on the conditions of the voice disorders database. To find association rules over transactional medical database first we use apriori algorithm for frequent item set mining. These two initial steps of analysis will help to create the medical knowledgebase. The ultimate goal is to build a model, which can improve the way to read and interpret the existing data in medical database and future data as well.

TEXT CLASSIFICATION USING MODIFIED MULTI CLASS ASSOCIATION RULE

This paper presents text classification using a modified Multi Class Association Rule Method. The method is based on Associative Classification which combines classification with association rule discovery. Although previous work proved that Associative Classification produces better classification accuracy compared to typical classifiers, the study on applying Associative Classification to solve text classification problem are limited due to the common problem of high dimensionality of text data and this will consequently results in exponential number of generated classification rules. To overcome this problem the modified Multi-Class Association Rule Method was enhanced in two stages. In stage one the frequent pattern are represented using a proposed vertical data format to reduce the text dimensionality problem and in stage two the generated rule was pruned using a proposed Partial Rule Match to reduce the number of generated rules. The proposed method was tested on a text classification problem and the result shows that it performed better than the existing method in terms of classification accuracy and number of generated rules.

An enhanced incremental association rule discovery with a lower minimum support

In the real world of data, a new set of data has been being inserted into the existing database. Thus, the rule maintenance of association rule discovery in large databases is an important problem. Every time the new data set is appended to an original database, the old rule may probably be valid or invalid. This paper proposed the approach to calculate the lower minimum support for collecting the expected frequent itemsets. The concept idea is applying the normal approximation to the binomial theory. This proposed idea can reduce a process of calculating probability value for all itemsets that are unnecessary. In addition, the confidence interval is also applied to ensure that the collection of expected frequent itemsets is properly kept.

Finding Association Rules in Medical Datasets

Finding Association Rules in Medical Datasets

A Comparative Study of Pattern Recognition Algorithms on Sales Data

In the realm of Data Mining looking for patterns and association rules is a very critical task and has been widely studied in the past years. There exist several data mining algorithms to find Association Rules in given datasets. One of the most popular and widely used algorithm is the Apriori algorithm to find patterns and itemsets in huge datasets and getting the association rules between them. This is done to gather knowledge from otherwise unsuspecting and random data. The Fp-Growth algorithm is similarly a different algorithm which uses an extended frequent pattern prefix-tree data structure for storing critical data after compression about frequent pairs. In this paper we do a comparative analysis of the 2 most popular pattern recognition algorithms and their performance on sales data of a college canteen sales transnational database where each record consists of items purchased by customer. General Terms Pattern Recognition, Data Mining, Apriori, Fp-Growth, Frequent Itemsets, Item-pair.

A Comparative Study of Tree-Based and Apriori-Based Approaches for Incremental Data Mining

Association rule mining is an iterative and interactive process of discovering valid, novel, useful, understandable and hidden associations from the massive database. The Colossal databases require powerful and intelligent tools for analysis and discovery of frequent patterns and association rules. Several researchers have proposed the many algorithms for generating item sets and association rules for discovery of frequent patterns, and minning of the association rules. These proposals are validated on static data. A dynamic database may introduce some new association rules, which may be interesting and helpful in taking better business decisions. In association rule mining, the validation of performance and cost of the existing algorithms on incremental data are less explored. Hence, there is a strong need of comprehensive study and in-depth analysis of the existing proposals of association rule mining. In this paper, the existing tree-based algorithms for incremental data mining are presented and compared on the baisis of number of scans, structure, size and type of database. It is concluded that the Can-Tree approach dominates the other algorithms such as FP-Tree, FUFP-Tree, FELINE Alorithm with CATS-Tree etc.This study also highlights some hot issues and future research directions. This study also points out that there is a strong need for devising an efficient and new algorithm for incremental data mining.

Computing exact permutation p-values for association rules

Association rule mining is an important task in the field of data mining, and many efficient algorithms have been proposed to address this problem. However, a large portion of the rules reported by these algorithms just satisfy the user-defined constraints purely by accident, and those that are not statistically meaningful should be filtered out through statistical significance testing. In the context of association rule discovery, the permutation-based approach can achieve better performance than other competitive methods, although several drawbacks of this effective approach narrow its usability. In this paper, we provide an analysis of these disadvantages and propose an algorithm called Exact Permutation p-values for Association Rules (EPAR) to calculate the exact p-values of all tested rules. Experiments on different types of data sets demonstrate that EPAR can successfully alleviate the disadvantages and outperform the direct permutation-based method over several performance measures.

Effect of Drug-Drug Interactions by Association Rule Discovery among the Elderly

Objective: With the decline of the bodily function, the elderly increased polypharmacy frequency. The results lead to the danger for duplicate drug and drug interactions (DDIs). This study aims to analyze the effect of DDIs of the elderly in Taiwan. Methods: The association rules and hypothesis testing technology were proposed to analyze the impact of drug-drug interactions, using the real cases in Taiwan Teaching Hospital. Results: From the analysis, shown three results. Firstly, four out of 100 patients have the threat of drug interactions. Secondly, the proportion of drug-drug interactions is highest in the cardiovascular medicine. Thirdly, the number of drug-drug interaction for elderly is 2.01 times of the other people. Conclusion: I suggested that the patient should be reminded to describe their medications for medical staffs spontaneously to help doctors for reducing the risk of drug interactions and drug wastage.

Data Mining Framework for Generating Sales Decision Making Information Using Association Rules

The rapid technological development in the field of information and communication technology (ICT) has enabled the databases of super shops to be organized under a countrywide sales decision making network to develop intelligent business systems by generating enriched business policies. This paper presents a data mining framework for generating sales decision making information from sales data using association rules generated from valid user input item set with respect to the sales data under analysis. The proposed framework includes super shop’s raw database storing sales data collected through sales application systems at different Point of Sale (POS) terminals. Apriori algorithm is famous for association rule discovery from the transactional database. The proposed technique using customized association rule generation and analysis checks the input items with sales data for validation of the input items. The support and confidence of each rule are computed. Sales decision making information about input items is generated by analyzing each of the generated association rules, which can be used to improve sales decision making policy to attract customers in order to increase sales. It is hoped that this approach for generating sales decision making information by analyzing sales data using association rules is more specific decision and application oriented as the business decision makers are not usually interested to all of the items of the sales database for making a specific sales decision.

Mining frequent biological sequences based on bitmap without candidate sequence generation

Biological sequences carry a lot of important genetic information of organisms. Furthermore, there is an inheritance law related to protein function and structure which is useful for applications such as disease prediction. Frequent sequence mining is a core technique for association rule discovery, but existing algorithms suffer from low efficiency or poor error rate because biological sequences differ from general sequences with more characteristics. In this paper, an algorithm for mining Frequent Biological Sequence based on Bitmap, FBSB, is proposed. FBSB uses bitmaps as the simple data structure and transforms each row into a quicksort list QS-list for sequence growth. For the continuity and accuracy requirement of biological sequence mining, tested sequences used during the mining process of FBSB are real ones instead of generated candidates, and all the frequent sequences can be mined without any errors. Comparing with other algorithms, the experimental results show that FBSB can achieve a better performance on both run time and scalability.

Discovery of the D-basis in binary tables based on hypergraph dualization

Discovery of (strong) association rules, or implications, is an important task in data management, and it finds application in artificial intelligence, data mining and the semantic web. We introduce a novel approach for the discovery of a specific set of implications, called the D-basis, that provides a representation for a reduced binary table, based on the structure of its Galois lattice. At the core of the method are the D-relation defined in a lattice theoretic framework and the hypergraph dualization algorithm that allows us to effectively produce the set of transversals for a given Sperner hypergraph. The latter algorithm, first developed by specialists from Rutgers Center for Operations Research, has already found numerous applications in solving optimization problems in data base theory, artificial intelligence and game theory. One application of the method is for analysis of gene expression data related to a particular phenotypic variable, and some initial testing is done for data provided by the University of Hawaii Cancer Center.

FIR: An Effective Scheme for Extracting Useful Metadata from Social Media.

Recently, the use of social media for health information exchange is expanding among patients, physicians, and other health care professionals. In medical areas, social media allows non-experts to access, interpret, and generate medical information for their own care and the care of others. Researchers paid much attention on social media in medical educations, patient-pharmacist communications, adverse drug reactions detection, impacts of social media on medicine and healthcare, and so on. However, relatively few papers discuss how to extract useful knowledge from a huge amount of textual comments in social media effectively. Therefore, this study aims to propose a Fuzzy adaptive resonance theory network based Information Retrieval (FIR) scheme by combining Fuzzy adaptive resonance theory (ART) network, Latent Semantic Indexing (LSI), and association rules (AR) discovery to extract knowledge from social media. In our FIR scheme, Fuzzy ART network firstly has been employed to segment comments. Next, for each customer segment, we use LSI technique to retrieve important keywords. Then, in order to make the extracted keywords understandable, association rules mining is presented to organize these extracted keywords to build metadata. These extracted useful voices of customers will be transformed into design needs by using Quality Function Deployment (QFD) for further decision making. Unlike conventional information retrieval techniques which acquire too many keywords to get key points, our FIR scheme can extract understandable metadata from social media.

Knowledge discovery and sequence-based prediction of pandemic influenza using an integrated classification and association rule mining (CBA) algorithm

Pandemic influenza is a major concern worldwide. Availability of advanced technologies and the nucleotide sequences of a large number of pandemic and non-pandemic influenza viruses in 2009 provide a great opportunity to investigate the underlying rules of pandemic induction through data mining tools. Here, for the first time, an integrated classification and association rule mining algorithm (CBA) was used to discover the rules underpinning alteration of non-pandemic sequences to pandemic ones. We hypothesized that the extracted rules can lead to the development of an efficient expert system for prediction of influenza pandemics. To this end, we used a large dataset containing 5373 HA (hemagglutinin) segments of the 2009 H1N1 pandemic and non-pandemic influenza sequences. The analysis was carried out for both nucleotide and protein sequences. We found a number of new rules which potentially present the undiscovered antigenic sites at influenza structure. At the nucleotide level, alteration of thymine (T) at position 260 was the key discriminating feature in distinguishing non-pandemic from pandemic sequences. At the protein level, rules including I233K, M334L were the differentiating features. CBA efficiently classifies pandemic and non-pandemic sequences with high accuracy at both the nucleotide and protein level. Finding hotspots in influenza sequences is a significant finding as they represent the regions with low antibody reactivity. We argue that the virus breaks host immunity response by mutation at these spots. Based on the discovered rules, we developed the software, “Prediction of Pandemic Influenza” for discrimination of pandemic from non-pandemic sequences. This study opens a new vista in discovery of association rules between mutation points during evolution of pandemic influenza.