Frequent Pattern Mining Research Articles

Discovery of Frequent Patterns of Episodes Within a Time Window for Alarm Management Systems

The sequential pattern mining field is expanding through numerous researches and has a large number of applications such as language processing, alarms management and event management on a broader scale. Its use began with processing items baskets to learn patterns and have a directed marketing strategy but it is generalized to telecommunication alarms management with several works. Our work is in line with this, as it tries to locate patterns and identify them to make predictive statements about certain patterns. It is axed around providing a way to break sequences into episodes and assigning them a value of confidence and support, more precisely in the discovery of frequent patterns of episodes within a time window. Experimental results have shown the effectiveness of our sequential pattern mining approach and its adaptability to alarm management and analytics.

HDFS-based parallel and scalable pattern mining using clouds for incremental data

Increased usage of internet led to the migration of large amount of data to the cloud environment which uses Hadoop and map reduce framework for managing various mining applications in distributed environment. Earlier research activity in distributed mining comprises of solving complex problems using distributed computational techniques and new algorithmic designs. But as the nature of the data and user requirement becomes more complex and demanding, the existing distributed algorithms fails in multiple aspects. In our work, a new distributed frequent pattern algorithm, named Hadoop-based parallel frequent pattern mining (HPFP) has been proposed to optimally utilise the clusters efficiently and mine repeated patterns from large databases very effectively. The empirical evaluation shows that HPFP algorithm improves the performance of mining operation by increasing the level of parallelism and execution efficacy. HPFP achieves complete parallelism and delivers superior performance to become an efficient algorithm in HDFS, than existing distributed pattern mining algorithms.

Low-cohesion differential privacy protection for industrial Internet

Due to the increasing intelligence of data acquisition and analysis in cyber physical systems (CPSs) and the emergence of various transmission vulnerabilities, this paper proposes a differential privacy protection method for frequent pattern mining in view of the application-level privacy protection requirements of industrial interconnected systems. This method designs a low-cohesion algorithm to realize differential privacy protection. In the implementation of differential privacy protection, Top-k frequent mode method is introduced, which combines the factors of index mechanism and low cohesive weight of each mode, and the original support of each selected mode is disturbed by Laplacian noise. It achieves a balance between privacy protection and utility, guarantees the trust of all parties in CPS and provides an effective solution to the problem of privacy protection in industrial Internet systems.

HDFS-based parallel and scalable pattern mining using clouds for incremental data

Increased usage of internet led to the migration of large amount of data to the cloud environment which uses Hadoop and map reduce framework for managing various mining applications in distributed environment. Earlier research activity in distributed mining comprises of solving complex problems using distributed computational techniques and new algorithmic designs. But as the nature of the data and user requirement becomes more complex and demanding, the existing distributed algorithms fails in multiple aspects. In our work, a new distributed frequent pattern algorithm, named Hadoop-based parallel frequent pattern mining (HPFP) has been proposed to optimally utilise the clusters efficiently and mine repeated patterns from large databases very effectively. The empirical evaluation shows that HPFP algorithm improves the performance of mining operation by increasing the level of parallelism and execution efficacy. HPFP achieves complete parallelism and delivers superior performance to become an efficient algorithm in HDFS, than existing distributed pattern mining algorithms.

A Combination of Frequent Pattern Mining and Graph Traversal Approaches for Aspect Elicitation in Customer Reviews

Due to the remarkable increase in e-commerce transactions, people try to have an appropriate choice of purchase through considering other people's reflected experience in product's or service's reviews. Automatic analysis of such corpus requires enhanced developed algorithms based on natural language processing and opinion mining. Moreover, the linguistic differences make extending existing algorithms from one language to another challenging and in some cases impossible. Opinion mining focuses on different subjects of review analysis such as spam detection, aspect elicitation and polarity allocation. In this article, we focus on detection of explicit aspect and propose a methodology to overcome some difficult and problematic aspect compounds in the form of multi- words format in Persian language. Our approach proposes the construction of a directed weighted graph (ADG structure) based on some yielded information from FP-Growth frequent pattern identification algorithm on our corpus of Persian sentence. Traversing some special paths within the ADG graph according to our developed rules could lead us to the extraction of problematic multi-word aspects. We utilize Neo4j NoSQL graph database environment and its Cypher query language in order to create the ADG graph and access the desired paths that reflects our developed rules on the ADG structure which lead us to extract the multi-word aspects. The evaluation of our methodology with the existing approaches on the issue of aspect derivation in Persian language including ELDA, SAM, an MMI-based and an LRT-based algorithms indicates the robustness of our approach.

Alarm Correlation to improve industrial fault management

Abstract Actual alarm systems used in manufacturing applications lacks explanation and indication of the root causes, which results in a poor decision making. In addition, manufacturing systems are more and more complex, so relaying on human operators for alarm information management becomes impossible. For this, a computerized tool to support human operators (i.E., decision support for information management system) is needed and would increase analytical capability for alarm analysis. To this effect, we introduce in this paper an autonomous data mining method to search historical alarm logs for the correlations that can represent causal relationships, which can support alarm management and system improvement. We investigate the use of Frequent Pattern Mining algorithm, an enumeration-tree based approach, for extracting relationships and automatically detect correlation between industrial alarms. Due to the time indexation of the alarm events, we adapt the algorithm in order to take into account the duration between alarms when extracting the itemsets. Filtered rules where evaluated according to the Minimmum support & Confidence framework. Obtained results show that FPM algorithm can derive very useful knowledge on system behaviour allowing the identification of alarm subsequence with the corresponding root cause.

Research on Distributed Vertical Frequent Pattern Mining Method Based on Metadata Integration

In order to better integrate the information industry into people’s life and work, and even in social development, how to mine data has become a hot issue. Metadata is a kind of data about data. Mining metadata helps data application and storage. Therefore, it is very important to find an efficient and intelligent data mining method. This paper introduces the metadata and its integration technology. On this basis, it introduces the distributed vertical frequent mode, and introduces its use in the process of mining metadata, providing a new working idea for the staff engaged in related industries.

Weighted Based Frequent and Infrequent Pattern Mining Model for Real-time E-Commerce Databases

Weighted Based Frequent and Infrequent Pattern Mining Model for Real-time E-Commerce Databases

An improved Frequent Pattern Mining in Sustainable Learning Practice using Generalized Association Rules

This research focuses on mining the frequent patterns occurred in the given Datasets by Generalization of Association Rules. Frequent pattern mining is a significant as well as interesting problem in the research filed of Data Mining. Building of frequent pattern tree (FP tree), frequent pattern growth (FP growth) and association rule generation are implemented to find the repeated patterns of data. FP tree Construction Algorithm is mainly used to compact a vast database into a extremely compressed, seems to very tiny data structure; hence eliminates the repeated scanning of database. The role of FP growth algorithm is to discover the frequent patterns with FP tree structure and construct the generalized association rules corresponding to the learning data. FP growth algorithm obtained best results as compared with conventional Apriori algorithm. This research provides some practical real time applications pertaining data mining techniques in the field of learning, education, marketing, finance and so on.

Efficient Mining of Interest Patterns on Click Stream Data

Nowadays, large amount of data is generated daily in e-commerce applications as click stream data. Because of the availability of this tremendous amount of data analyzing the user browsing behaviour and finding frequent navigation patterns of different web pages accessed by web users is an key element for retailers to optimize the website and personalized the web services of different e-commerce websites. User browsing behaviour is evaluated based on user interests on web pages or products. There are different parameters are considered while analyzing the click stream data for calculating frequent navigation patterns and context based customer behaviour in online data bases. In this paper we developed different models for optimizing and personalizing web service and sequential frequent patterns using the parameters: browsing path, frequently visited web pages, time duration of web pages and user interest. These novel models uses the parameters and applied on click stream data to optimize the web pages and improve the personalized recommendation.

A roadside unit deployment framework for enhancing transportation services in Maghrebian cities

SummaryRoadside units (RSUs) have a crucial role in maintaining vehicular ad hoc networks (VANETs) connectivity and coverage, especially, for applications gathering or disseminating nonsafety information. In big cities with complex road network topology, a huge number of costly RSUs must be deployed to collect data gathered by all moving vehicles. In this respect, several research works focusing on RSUs deployment have been proposed. The thriving challenge would be to (1) reduce the deployment cost by minimizing as far as possible the number of used RSUs and (2) to maximize the coverage ratio. In this paper, we introduce a spatiotemporal RSU deployment framework including three methods, namely, SPaCov, SPaCov+, and HeSPic. SPaCov starts by mining frequent mobility patterns of moving vehicles from their trajectories; then, it computes the best RSU locations that cover the extracted patterns. Nonetheless, SPaCov+ extracts the frequent mobility patterns as well as the rare ones to enhance the coverage ratio. HeSPiC is a budget‐constrained spatiotemporal coverage method that aims to maximize the coverage ratio subject to a budget constraint, which is defined in terms of RSU number. Performed simulations highlight the efficiency and the effectiveness of the proposed RSU deployment framework in terms of coverage ratio, deployment cost, network latency and overhead.

WMFP-Outlier: An Efficient Maximal Frequent-Pattern-Based Outlier Detection Approach for Weighted Data Streams

Since outliers are the major factors that affect accuracy in data science, many outlier detection approaches have been proposed for effectively identifying the implicit outliers from static datasets, thereby improving the reliability of the data. In recent years, data streams have been the main form of data, and the data elements in a data stream are not always of equal importance. However, the existing outlier detection approaches do not consider the weight conditions; hence, these methods are not suitable for processing weighted data streams. In addition, the traditional pattern-based outlier detection approaches incur a high time cost in the outlier detection phase. Aiming at overcoming these problems, this paper proposes a two-phase pattern-based outlier detection approach, namely, WMFP-Outlier, for effectively detecting the implicit outliers from a weighted data stream, in which the maximal frequent patterns are used instead of the frequent patterns to accelerate the process of outlier detection. In the process of maximal frequent-pattern mining, the anti-monotonicity property and MFP-array structure are used to accelerate the mining operation. In the process of outlier detection, three deviation indices are designed for measuring the degree of abnormality of each transaction, and the transactions with the highest degrees of abnormality are judged as outliers. Last, several experimental studies are conducted on a synthetic dataset to evaluate the performance of the proposed WMFP-Outlier approach. The results demonstrate that the accuracy of the WMFP-Outlier approach is higher compared to the existing pattern-based outlier detection approaches, and the time cost of the outlier detection phase of WMFP-Outlier is lower than those of the other four compared pattern-based outlier detection approaches.

Recognizing patterns of student\u2019s modeling behaviour patterns via process mining

Evaluating student learning effect plays an essential role in education, which is typically done by assessing student’s final deliverables. However, the student’s learning process has not been properly explored in the past.In this paper, we propose an interactive student learning effect evaluation framework which focuses on in-process learning effect evaluation. In particular, our proposal analyzes students modeling assignment based on their operation records by using techniques of frequent sequential pattern mining, user behavior analysis, feature engineering, and process mining. A comprehensive online modeling platform has been developed to enable data collection. We have carried out a case study, in which we applied our approach to a real teaching scenario, consisting of student online modeling behavior data collected from 24 students majoring in computer science. We also associate our process mining results with the numeric evaluation values. The preliminary result of case analysis has shown good potential to mine student modeling patterns and interpret their behaviors, contributing to student learning effect evaluation.

Penerapan Data Mining Apriori Pada Persediaan Obat (Studi Kasus Apotek Rafif Farma Medan)

<span><em>The importance of inventory systems at a pharmacy and the type of goods which</em><br /><span><em>are a top priority that must be in stock. It is useful to anticipate the void stuff. Due to the</em><br /><span><em>lack of inventory may affect customer service and asset to the pharmacy. Therefore, this</em><br /><span><em>study was conducted to help resolve those problems by designing a data mining</em><br /><span><em>application that serves to predict sales of the drug is needed most knowable a priori</em><br /><span><em>algorithm with the help of Tools Tanagra. One of the interesting association analysis</em><br /><span><em>phase analysis algorithm that generates a high frequency patterns (frequent pattern</em><br /><span><em>mining).</em><br /><span><em>Keywords: Data Mining, Apriori Algorithm, Association Rule</em></span></span></span></span></span></span></span></span><br /><br class="Apple-interchange-newline" /></span>

Various Research Opportunities in High Utility Itemset Mining

Pattern mining is a technique, which discovers interesting, hidden, unpredicted and useful patterns of data from the database. Most of the research work in pattern mining has been focused on the traditional way of Frequent Itemset Mining (FIM) and Association Rule Mining (ARM) for pattern-discovery. Patterns in frequent itemset mining are based on the occurrence frequency of items. Although frequent pattern mining is useful, the assumption that ‘frequent patterns are interesting,’ doesn’t hold for numerous applications. High Utility Itemset Mining (UIM) overcomes this limitation of frequent itemset mining. The aim of HUIM is to find the patterns based on a utility function where the utility can be measured in terms of revenue, profit, weight, frequency, interestingness or time spent on some webpage, etc. Mining patterns with high utility can be seen as a generalization of FIM where the transaction database is the input and every item is having a utility factor representing its importance and might have non-binary quantities in the transactions. This paper surveys various recent advances and research opportunities in the field of high utility itemset mining.

A New Pattern Mining Algorithm for Analytics of Real-Time Internet of Things Data

The rise of IoT Real time data has led to new demands for mining systems to learn complex models with millions to billions of parameters, which promise adequate capacity to digest massive datasets and offer powerful predictive analytics. To support Big Data mining, high-performance powerful computing platforms are required, which impose regular designs to unleash the full power of the Big Data. Pattern mining poses a lot of interesting research problems and there are many areas that are still not well understood. The specifically very elementary challenges are to understand the meaningful data from the junk data that anticipated into the internet, refer as “Smart Data”. Eighty-five percent of the entire data are noisy or meaningless. It is a very tough often assigned to verify and separate to refine the data from the noisy junk. Researchers’ are proposing an algorithm of distributed pattern mining to give some sort of solution of the heterogeneity, scaling and hidden Big Data problems. The algorithm has evaluated in parameters like cost, speed, space and overhead. Researchers’ used IoT as the source of Big Data that generates heterogeneous Big Data. In this paper, we are representing the results of all tests proved that; the new method gives accurate results and valid outputs based on verifying them with the results of the other valid methods. Also, the results show that, the new method can handle the big datasets and decides the frequent pattern and produces the associate rule sets faster than that of the conventional methods and less amount of memory storage for processing. Overall the new method has a challenging performance as regard the memory storage and the speed of processing as compared to the conventional methods of frequent pattern mining like Apriori and FP-Growth techniques.

이진 논리 연산을 활용한 하이 BHUI: 유틸리티 마이닝 기법에 대한 연구

Pattern mining finding frequent patterns are widely used to search useful information from large scale data. General pattern mining, however, have the problem of not applying the characteristics of the individual items because all items have the same importance. High utility itemset mining that solves this problem retrieves meaningful information by applying a relative weight to each item. In this paper, we propose a new high utility itemset mining algorithm based on bitwise operations. The bitwise operation predetermines the likelihood of joining two itemsets to reduce the join operation. Since the join operation is the most time consuming operation in the high utility itemset mining, the proposed algorithm significantly reduces the mining time. We analyzed the performance of the proposed algorithm with execution time and the number of generated utility itemsets and compared the performance with HUI-Miner. Our experimental results show that the proposed algorithm outperforms HUI-Miner.

Efficient approach for incremental weighted erasable pattern mining with list structure

Erasable pattern mining is one of the important fields of frequent pattern mining. It diagnoses and solves the economic problems that arise in the manufacturing industry. The real-world database is continually accumulated over time, and each item has a different importance. Therefore, if we use conventional erasable pattern mining without considering the characteristics of the real-world database, less meaningful patterns can be extracted. Also, when mining a real-world database, the algorithm must be able to process operations quickly and efficiently. In this paper, in order to meet these requirements, we propose an algorithm which is implemented as a list structure for mining erasable patterns in an incremental database with weighted condition. Compared to existing state-of-the-art mining algorithms, the proposed algorithm performs pattern pruning by applying weighted condition to a dynamic database, so it extracts fewer candidate patterns and shows fast performance. We test our algorithms and the algorithms previously presented with various real datasets and synthetic datasets and obtained results such as run time, memory usage, scalability, and accuracy tests. By analyzing and comparing these experimental results, we show that the proposed algorithm has outstanding performance.

A text analytics framework for automated communication pattern analysis

Human-coding reliant conversation analysis methods are ineffective when analyzing large volumes of data. In this paper, we propose a text analytics framework for automated conversation pattern analysis. This framework first extracts speech acts (i.e., activities) from conversation logs, and then analyzes their flow through frequent pattern mining algorithms to reveal insightful communication patterns. Using a real-world data set collected from a customer service center, we demonstrate the usefulness of the framework for identifying patterns that are associated with service quality outcomes. Our work has implications for the design of communication policies and systems for customer service management.

Mining Rank Data

The problem of frequent pattern mining has been studied quite extensively for various types of data, including sets, sequences, and graphs. Somewhat surprisingly, another important type of data, namely rank data, has received very little attention in data mining so far. In this article, we therefore address the problem of mining rank data, that is, data in the form of rankings (total orders) of an underlying set of items. More specifically, two types of patterns are considered, namely frequent rankings and dependencies between such rankings in the form of association rules. Algorithms for mining frequent rankings and frequent closed rankings are proposed and tested experimentally, using both synthetic and real data.