Search

In this paper, we propose a new method that combines the parallelism of the Spark-based platform with fast frequent mining, called STB_Apriori. Previous research has shown that traditional frequent itemset mining algorithms have high overhead when faced with large datasets and high-dimensional data computation, and generate a large number of candidate itemsets; at the same time, when faced with diverse user requirements, they often generate very sparse and diverse data. In order to solve the problem of fast mining of massive data, our idea originates from the capability of Spark distributed computing and the common optimisation ideas in Apriori mining, by using the efficient operator BitSet to achieve transaction compression, bit storage and data manipulation by Boolean matrices, and at the same time by parallelising the processing and optimising the algorithmic logic to achieve fast and frequent mining. In experiments on real-world datasets, our model consistently outperforms five widely used methods by a significant margin on very large data and maintains its excellence in the remaining cases, proving its effectiveness on real-world tasks, while further analysis shows that increasing the number of distributed nodes also incrementally and continuously improves performance. [ABSTRACT FROM AUTHOR]

Knowledge graphs (KGs) are popularly used to develop several intelligent applications. Revealing valuable knowledge hidden in these graphs opened up a branch of research, known as KG reasoning, aiming at predicting the missing links. Some methods take advantage of external information such as entity description but at the cost of more computational complexity. Besides, most of the current techniques focus solely on local information in the KG. However, the learning process can utilise valuable global information in the entire graph. In this paper, we propose a Pattern-based Knowledge Graph Completion (PKGC) method that consists of three phases. The first phase utilizes multi-source information and expands the KG using entity description as external information with efficient Natural Language Processing (NLP) techniques. In the second phase, we mine frequent patterns from the expanded KG, extract connections between them and assign entities to the patterns that construct the abstraction layer. Based on the extracted patterns, connections, and entity assignments, a flow network is constructed on the abstraction layer in the third phase. We use global internal information, namely patterns, by adapting the minimum-cost circulation problem to the flow network. This way the links in a larger neighborhood are involved in the inference. We conducted experiments on the link prediction task and evaluated the training time on two benchmark datasets, WordNet and Freebase. Experiments have demonstrated that the proposed method is superior to the state-of-the-art methods and that pattern extraction is effective for knowledge graph completion tasks.

Community detection is a significant research area in social networks. Most methods use network topology, but combining it with user interactions improves accuracy. This paper proposes a robust method to identify communities based on the improved user interaction degree, the weighted quasi-local structural similarity measure, and the frequent pattern mining on user interactions. In the community creation phase, influential users are identified based on eigenvector centrality and users who interact with them the most are extracted based on frequent pattern mining. In the community expansion phase, we introduce a measure to calculate the degree of user interactions based on the local clustering coefficient improved by interactions between common neighbors. We present two strategies to expand the community. The first strategy, a direct connection, exists between a user outside and a user inside the community. Their similarity is calculated based on the combined measure of improved user interaction degree and user degrees. The second strategy is if two users do not have a direct connection, we consider their communication paths. Therefore, we present a similarity measure combining a quasi-local path-based measure and an improved user interaction degree. Analysis of Higgs Twitter and Flickr datasets using internal density, Normalized Mutual Information, and Adjusted Rand Index shows that this paper's method outperforms the other five community detection methods. Furthermore, our method has more robustness than other relevant methods. [ABSTRACT FROM AUTHOR]

The high utility frequent pattern mining algorithm mines more important frequent patterns from the data by using the importance degree information. On this basis, the high utility quantitative frequent pattern mining algorithm further explores the quantitative relationship between data items, and thus has become a popular research topic in the field of data mining. RHUQI-Miner is proposed to improve the performance and practicability of the algorithm. Firstly, the concept of related degree is proposed, the item related degree structure is constructed according to the related degree, and a pruning optimization strategy is given to find frequent patterns with higher related degree, reducing redundancy and invalid frequent patterns. Secondly, the fixed pattern length strategy is used to modify the utility information of the item in the mining process, so that the algorithm can control the length of the output frequent pattern according to the actual data situation, and further improve the performance and practicability of the algorithm. The experimental results show that RHUQI-Miner can effectively reduce the time and memory consumption in the mining process, which can provide data support for differentiated and precise maintenance strategies. [ABSTRACT FROM AUTHOR]

It is of great significance to improve the efficiency of railway production and operation by realizing the fault knowledge association through the efficient data mining algorithm. However, high utility quantitative frequent pattern mining algorithms in the field of data mining still suffer from the problems of low time-memory performance and are not easy to scale up. In the context of such needs, we propose a related degree-based frequent pattern mining algorithm, named Related High Utility Quantitative Item set Mining (RHUQI-Miner), to enable the effective mining of railway fault data. The algorithm constructs the item-related degree structure of fault data and gives a pruning optimization strategy to find frequent patterns with higher related degrees, reducing redundancy and invalid frequent patterns. Subsequently, it uses the fixed pattern length strategy to modify the utility information of the item in the mining process so that the algorithm can control the length of the output frequent pattern according to the actual data situation and further improve the performance and practicability of the algorithm. The experimental results on the real fault dataset show that RHUQI-Miner can effectively reduce the time and memory consumption in the mining process, thus providing data support for differentiated and precise maintenance strategies.

Traditional periodic pattern mining ignoreds the correlation and recency of the patterns, and therefore obtains some weakly correlated and recency patterns with limited practical value. To address the issue, this paper proposed novel periodic pattern mining methods based on recency and correlation constraints named CRPFP-BFS and CRPFP-DFS. By transforming a given database into a column-based structure CRPFP-List, CRPFP-BFS and CRPFP-DFS recursively mined its data using breadth-first and depth-first search, respectively. At the same time, CRPFP-BFS and CRPFP-DFS applied four pruning strategies based on support, period, recency and correlation to reduce the search space, thereby effectively discovering correlation recency periodic patterns. By comparing with the current state-of-the-art algorithms on dense and sparse datasets, and the experimental results show that the CRPFP-BFS and CRPFP-DFS have lower memory usage, higher operating efficiencies and good scalability. Among them, CRPFP-DFS is suitable for situations with strict memory requirements, and CRPFP-BFS performs more efficiently for long transaction sparse databases. [ABSTRACT FROM AUTHOR]

Recommender systems (RS) are substantial for online shopping or digital content services. However, due to some data characteristics or insufficient historical data, may encounter considerable difficulties impacting the quality of their recommendations. This study introduces the clustering-based frequent pattern mining framework for recommender systems (Clustering-based FPRS) - a novel RS constituting several recommendation strategies leveraging agglomerative clustering and FP-growth algorithms. The developed strategies combine the generated frequent itemsets with collaborative- and content-filtering methods to address the cold-start problem, which occurs whenever a new user or item enters the system. In such cases, the RS has limited information about the new user or object. Thus, the recommendations may be inaccurate. The experimental evaluation on several benchmark datasets showed that Clustering-based FPRS is superior to state-of-the-art and could effectively alleviate the cold-start problem.

It is always difficult to analyze the law of ship movement, especially the behavior of ship. In this paper, a frequent pattern mining method based on a large number of historical track clustering of ships is proposed to estimate the future behavior of ships, and the software implementation is presented. In this paper, a comprehensive similarity measurement method of track is proposed and the meaning of frequent pattern mining based on track clustering is introduced. Secondly, the adaptive transformation of the classical density clustering algorithm is carried out and the implementation method of the clustering algorithm based on comprehensive similarity is given. Then, the most similar cluster of virtual trunk track calculation is extracted, and the estimation results of current ship behavior are obtained by statistics. Finally, the software design and test results based on C/S architecture are given. Experimental results show that this method can describe the behavior of track association, and the behavior estimation results obtained by the software can assist the research and judgment.

In order to effectively solve the problem of acquiring knowledge from tactical wargaming data, an overall analysis framework is designed based on the standard process of data mining. The data is analyzed from four aspects: time, space, maneuver path and multi-operator behavior correlation. The behavioral characteristics of single operators at different stages and the spatial distribution of key points such as shooting points, hit points and hidden points, and the association rules of movement, shooting, and occupation between multiple operators are obtained. This will provide commanders with experience and knowledge, help them to quickly accumulate combat experience, and provide behavior rules and action modes for the development of wargaming AI, effectively improving its intelligent level. [ABSTRACT FROM AUTHOR]

A hypergraph is a complex data structure capable of expressing associations among any number of data entities. Overcoming the limitations of traditional graphs, hypergraphs are useful to model real-life problems. Frequent pattern mining is one of the most popular problems in data mining with a lot of applications. To the best of our knowledge, there exists no flexible pattern mining framework for hypergraph databases decomposing associations among data entities. In this article, we propose a flexible and complete framework for mining frequent patterns from a collection of hypergraphs. To discover more interesting patterns beyond the traditional frequent patterns, we propose frameworks for weighted and uncertain hypergraph mining also. We develop three algorithms for mining frequent, weighted, and uncertain hypergraph patterns efficiently by introducing a canonical labeling technique for isomorphic hypergraphs. Extensive experiments have been conducted on real-life hypergraph databases to show both the effectiveness and efficiency of our proposed frameworks and algorithms. [ABSTRACT FROM AUTHOR]

It is always difficult to analyze the law of ship movement, especially the behavior of ship. In this paper, a frequent pattern mining method based on a large number of historical track clustering of ships is proposed to estimate the future behavior of ships, and the software implementation is presented. In this paper, a comprehensive similarity measurement method of track is proposed and the meaning of frequent pattern mining based on track clustering is introduced. Secondly, the adaptive transformation of the classical density clustering algorithm is carried out and the implementation method of the clustering algorithm based on comprehensive similarity is given. Then, the most similar cluster of virtual trunk track calculation is extracted, and the estimation results of current ship behavior are obtained by statistics. Finally, the software design and test results based on C/S architecture are given. Experimental results show that this method can describe the behavior of track association, and the behavior estimation results obtained by the software can assist the research and judgment. [ABSTRACT FROM AUTHOR]

In recent years, community detection in social networks has become one of the most important research areas. One of the ways to community detection is to use interactions between users. There are different types of interactions in social networks, which, if used together with network topology, improve the precision of community identification. In this paper, a new method based on the combination of user interactions and network topology is proposed to community detection. In the community formation stage, the effective nodes are identified based on eigenvector centrality, and the primary communities around these nodes are formed based on frequent pattern mining. In the community expansion phase, small communities expand using modularity and the degree of interactions among users. To calculate the degree of interaction between users, a new measure based on the local clustering coefficient and interactions between common neighbors is proposed, which improves the accuracy of the degree of user interactions. Analysis of Higgs Twitter and Flickr datasets utilizing internal density metric, NMI and Omega demonstrates that the proposed method outperforms the other five community detection methods.

Periodic frequent-pattern mining (PFPM) is a vital knowledge discovery technique that identifies periodically occurring patterns in a temporal database. Although traditional PFPM algorithms have many applications, they often produce a large set of periodic-frequent patterns (PFPs) in a database. As a result, analyzing PFPs can be very time-consuming for users. Moreover, a large set of PFPs makes PFPM algorithms less efficient regarding runtime and memory consumption. This paper handles this problem by proposing a novel model of closed periodic-frequent patterns (CPFPs) found in databases. CPFPs are less expensive to mine because they represent a concise and lossless subset uniquely describing the entire set of PFPs. We also present an efficient depth-first search algorithm, called Closed Periodic-Frequent Pattern-Miner (CPFP-Miner), to discover the patterns. The proposed algorithm utilizes the weighted ordering of the patterns concept to reduce the patterns' search space. On the other hand, the current periodicity concept is also applied to prune aperiodic patterns from the search space. Extensive experiments on both real-world and synthetic databases demonstrate that the CPFP-Miner algorithm is efficient. It outperforms the state-of-the-art algorithms regarding runtime requirements, memory consumption, and energy consumption on several real-world and synthetic databases. Additionally, the scalability of the CPFP-Miner algorithm is demonstrated to be more effective and productive than the state-of-the-art algorithms. Finally, we present two case studies to show the functionality of the proposed patterns. [ABSTRACT FROM AUTHOR]

Data mining provides various frequent pattern mining methods to help business owners identify items with frequency or utility values greater than a particular threshold. Conversely, in the stock markets with a wide variety of stocks, a major challenge for buyers is to choose a set of items whose total price is less than their budget. In addition, many investors take into account the liquidity of the stock shares, which can return their capital at any time. However, the existing pattern mining techniques do not offer an appropriate solution to address this challenge. Therefore, the proposed framework tries to offer a set of stocks as a liquid purchasing portfolio by using pattern mining techniques and analyzing stock market data. First, by introducing the Ranking Tables, it explores stocks with the highest trading value that can be traded over time. Second, a new algorithm based on multivariate time series data is devised for mining frequent items that aims at selecting the maximum number of stocks whose buyers' power is greater than their sellers' power simultaneously. In addition, using stock prediction techniques, profitable stocks are identified and offered by considering the user's budget. The framework was compared with FTPM, FPTMwEVL, and MTS-FPM, frequent itemset mining algorithms on time-series data. In addition, it was evaluated in different aspects, like liquidity and profitability within the market. The results of the experiments indicate that this framework offers four times more liquidity as well as 60% more profitability over time compared to the whole market. [ABSTRACT FROM AUTHOR]

Trip recommendation for groups of tourists (TRGT) is a challenging task in tourism since many tourists travel in groups, inducing social interaction and bringing various social benefits. However, TRGT must address various real-life constraints such as limited time for touring, cost, etc. TRGT aims to design personalized tours that meet the preferences of all group members by addressing a variety of tourists' requirements that may sometimes result in conflicts and stress for the group members. TRGT should satisfy that both the preferences of group members need to be achieved as much as possible and the preferences of group members need to be achieved as evenly as possible. In this paper, we present a methodology for tackling the TRGT problem by reducing it to the Cooperative Location Set Cover Problem (CLSCP), formulated as an integer linear program. The CLSCP aims to select a group of facilities that can satisfy, in aggregate, all demand points. To tackle the CLSCP, we present a new method based on detecting frequent patterns. We also demonstrate the efficiency of the proposed methodology by presenting extensive experimental tests. [ABSTRACT FROM AUTHOR]

Mining frequent Time Intervals-Related Patterns (TIRPs) from series of symbolic time intervals offers a comprehensive framework for heterogeneous, multivariate temporal data analysis in various application domains. While gaining a growing interest in recent decades, the efficient mining of frequent TIRPs is still a high computational challenge which has also not yet been investigated in its full complexity. The majority of previous methods discover only the first instances of the TIRPs within each series of symbolic time intervals, whereas their re-occurring instances are ignored. This eventually results in an incomplete discovery of frequent TIRPs, a problem that lies also in the challenge of mining only the frequent closed TIRPs, which was only recently investigated for the first time. In this paper, we introduce TIRPClo—an efficient algorithm for the complete mining of either the entire set of frequent TIRPs, or only the frequent closed TIRPs. The algorithm proposes a non-ambiguous sequential representation of symbolic time intervals series through the intervals' end-points, as well as a memory-efficient index and a novel method for data projection, due to which it is the first algorithm to guarantee a complete discovery of frequent closed TIRPs. The experimental evaluation conducted on eleven real-world and four synthetic datasets demonstrates that TIRPClo is up to 10 times faster when mining the entire set of frequent TIRPs, and up to more than 100 times faster when mining only the frequent closed TIRPs compared to four state-of-the-art methods, while also reporting lower memory measurements. [ABSTRACT FROM AUTHOR]

This paper introduces a frequent pattern mining framework for recommender systems (FPRS) - a novel approach to address the items' cold-start problem. This difficulty occurs when a new item hits the system, and properly handling such a situation is one of the key success factors of any deployment. The article proposes several strategies to combine collaborative and content-based filtering methods with frequent items mining and agglomerative clustering techniques to mitigate the cold-start problem in recommender systems. The experiments evaluated the developed methods against several quality metrics on three benchmark datasets. The conducted study confirmed usefulness of FPRS in providing apt outcomes even for cold items. The presented solution can be integrated with many different approaches and further extended to make up a complete and standalone RS. [ABSTRACT FROM AUTHOR]

Frequent pattern mining (FPM) has played an important role in many graph domains, such as bioinformatics and social networks. In this paper, we focus on geo-social graphs, a kind of social network augmented by geographical information. However, in addition to the exponential time complexity of the problem, we face the challenge of efficient subgraph retrieval since we are interested in patterns in a specific region in such a network. For this reason, we formulate the top- $k$ FPM problem in large geo-social networks. Specifically, we devise a novel framework for subgraph retrieval and FPM mining with a series of optimizations. First, we propose a neighboring-aware R-tree (NaR-Tree) index structure to alleviate the challenge of retrieving subgraphs from a large graph. NaR-Tree is a variant of R-tree in which each nonleaf tree node further maintains some edge statistics information for the rectangle related to it. Second, we define the concept of minimum image-based support of edges (MNIE). With the help of the NaR-Tree and MNIE-based pattern extension approach, a mining algorithm that addresses the problem of exponential candidate patterns is proposed. We also present a lazy retrieval strategy to reduce the frequency of subgraph retrieval. Finally, we adopt an edge sampling approach to further accelerate the mining process. Extensive experiments on real-world and synthesized datasets are conducted to demonstrate the effectiveness and efficiency of our solution.

As a hot research topic in urban geography, spatiotemporal interaction analysis has been used to detect the hotspot mobility patterns of crowds and urban structures based on the origin-destination (OD) flow data, which provide useful information for urban planning and traffic management applications. However, existing methods mainly focus on the detection of explicit spatial interaction patterns (such as spatial flow clusters) in OD flow data, with less attention to the discovery of underlying crowd travel demands. Therefore, this paper proposes a framework to discover the crowd travel demands by associating the dynamic spatiotemporal interaction patterns and the contextual semantic features of the geographical environment. With urban functional zones (UFZs) as the basic units of human mobility in urban spaces, this paper gives a case study in Wuhan, China, to detect and interpret the human mobility patterns based on the characteristics of spatiotemporal interaction between UFZs. Firstly, we build the spatiotemporal interaction matrix based on the OD flows of different UFZs and analyze the characteristics of the interaction matrix. Then, hotspot poles, defined as the local areas where people gather significantly, are extracted using the Gi-statistic-based spatial hotspot detection algorithm. Next, we develop a frequent interaction pattern mining method to detect the frequent interaction patterns of the hotspot poles. Finally, based on the detected frequent interaction patterns, we discover the travel demands of crowds with semantic features of corresponding urban functional zones. The characteristics of crowd travel distance and travel time are further discussed. Experiments with floating car data, road networks, and POIs in Wuhan were conducted, and results show that the underlying travel demands can be better discovered and interpreted by the proposed framework and methods in this paper. This study helps to understand the characteristics of human movement and can provide support for applications such as urban planning and facility optimization. [ABSTRACT FROM AUTHOR]

This paper proposes an integrated system to control ramps and adjust variable speed limits. It includes three essential modules to predict the starting time of congestion and a fuzzy controller to determine the parameters and a model predictive control. An Apriori algorithm that is a powerful tool for frequent pattern mining is used in the first module. The proposed system is neither sensitive to the traffic distribution nor computationally intensive. Two traffic simulators of Aimsun and CTMSIM are applied to validate the results. Compared with the most recent algorithms, including Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM), this system improves prediction accuracy up to 2.63%. The results of ramp metering and variable speed limit subsystems are also promising. The embedded controller shows 0.6% and 4% overall and rush hour improvement in the total travel time.

The frequent conceptual links is a descriptive data mining task which aims at describing a social network in term of the most connected type of nodes. This is done by grouping nodes into clusters or groups according to their attributes and checking the number of links between the nodes of each couple of groups, if this number is greater than a predefined threshold, the set of links is referred to as a frequent conceptual link (FCL). Although relatively recent, this task has received a number of research, chiefly in order to optimize the exploration of the search process. Indeed, the problem is defined as NP-hard, where the search process depends on the size of the network, the number of attributes and the set of their possible values whose combination can explode quickly. In this paper, we propose a new algorithm for mining the frequent conceptual links in a social network based on the technique of the branch and bound. In addition to defining an upper bound for the potential patterns in the search space, the algorithm implements other techniques which improve significantly the performance of the search process and allows to fix the shortcomings constated in the previous implementations. [ABSTRACT FROM AUTHOR]

Call Detail Record (CDR) is the detailed record of all the telephonic calls that pass through a telephone exchange or any other telecommunications equipment. It contains temporal and spatial data, and can also convey other information that would be helpful to the user. Large numbers of vehicles on roads creates substantial traffic, which makes it very difficult to maintain safety and control traffic especially in the urban areas. Several works were carried out in the past to estimate the traffic density. However, they were inappropriate and quite expensive, owing to the dynamics of the traffic flow. This paper proposes the use of CDR data to find the high traffic density zones (HTDZs). For prediction purpose, we mine the frequent patterns from CDR data to find the co-occurrence of the position associated with a mobile user. In addition, Recurrent neural Networks (RNN) using LSTM (Long Short-term memory) are used for the time series prediction. The proposed system helps the whole public not only the registered users by decreasing the accident rates. Statistical performance evaluation integrated with time series causality is done for the proposed system. The proposed system is evaluated over standard data sets and an accuracy of 96% is achieved and a root mean square value was obtained as 3.84 during prediction. [ABSTRACT FROM AUTHOR]

Mining frequent patterns with single minimum support is quite unpredictable as it may miss rare itemsets depending upon the high value of minimum threshold. The phenomenon is commonly known as "rare item problem" dilemma. In this connection, frequent pattern mining under multiple minimum supports (MMS) is considered as an adequate solution of the problem. Existing approaches in this domain require user-defined minimum item support (MIS), and consider only the support in recognizing the importance of the items and ignore other factors including the inter-item relationships. The generation of huge number of itemsets including the itemsets with low associability or high dissociation is another problem in the existing MMS-based approaches. Keeping view of the above issues, this paper introduces AutoMMS-FPM, an automated MMS specification approach for frequent pattern mining with partial devoid of dissociation. The proposed approach not only takes account of support but also some other factors such as influence (inf), dissociation (d) and length importance factor (ρ) to boost up the mining process. Influence refers to the inter-item relationship and it is formulated based on the concept of degree centrality of the corresponding item network. MIS for the items are calculated automatically using MBS and RIF factors where the first one relates to the item support and the second one relates to the item influence. Experimental results on both of the synthetic and real datasets show that the proposed algorithm outperforms than existing approaches in terms of number of generated itemsets, run time, memory usages, dissociation and scalability. [ABSTRACT FROM AUTHOR]

A recommender system is an approach performed by e-commerce for increasing smooth users' experience. Sequential pattern mining is a technique of data mining used to identify the co-occurrence relationships by taking into account the order of transactions. This work will present the implementation of sequence pattern mining for recommender systems within the domain of e-commerce. This work will execute the Systolic tree algorithm for mining the frequent patterns to yield feasible rules for the recommender system. The feature selection's objective is to pick a feature subset having the least feature similarity as well as highest relevancy with the target class. This will mitigate the feature vector's dimensionality by eliminating redundant, irrelevant, or noisy data. This work presents a new hybrid recommender system based on optimized feature selection and systolic tree. The features were extracted using Term Frequency-Inverse Document Frequency (TF-IDF), feature selection with the utilization of River Formation Dynamics (RFD), and the Particle Swarm Optimization (PSO) algorithm. The systolic tree is used for pattern mining, and based on this, the recommendations are given. The proposed methods were evaluated using the MovieLens dataset, and the experimental outcomes confirmed the efficiency of the techniques. It was observed that the RFD feature selection with systolic tree frequent pattern mining with collaborative filtering, the precision of 0.89 was achieved. [ABSTRACT FROM AUTHOR]

Frequent itemset mining is extensively employed as an essential data mining technique. Nevertheless, as the data size grows, the applicability of this method decreases owing to the relatively poor performance of the existing methods. Though numerous efficient sequential frequent itemset mining methods have been developed, the performance that can be achieved is clearly limited by the fact that they exploit only one thread. To overcome these limitations, a number of parallel methods using multi-core central processing units (CPUs), multiple machines or many-core graphic processing units (GPU) have been proposed. However, these methods are relatively slow in performance and have low scalability, mainly owing to large memory requirements for intermediate data, significant disk I/Os, and heavy computation. In this study, to resolve the aforementioned problems, we propose ${\mathsf {SGMiner}}$ , which is a new, fast, and scalable GPU- and disk-based method on a single machine equipped with multiple graphic processing units (GPUs) and multiple solid-state drives (SSDs) for extracting frequent patterns. It is based on an algorithm similar to the Apriori algorithm and neither has intermediate data nor large disk I/O overheads owing to its exploitation of SSDs. Moreover, we propose storing transaction databases, namely bitmap transaction chunks, in SSDs, streaming the chunks to GPU device memory via the main memory with reduced I/O overhead, and performing fast support counting with GPUs based on the chunks. In addition, when exploiting multiple GPUs and SSDs, it proposes a concept of replicating bitmap transaction chunks stored in SSDs to GPUs in a streaming fashion. This could allow an almost equal workload to be distributed evenly across multiple GPUs with reduced I/O overheads. The experiments we conducted demonstrate that ${\mathsf {SGMiner}}$ outperforms the existing methods in terms of scalability and performance with enhanced robustness.

BackgroundConcerns about the role of chronically used medications in the clinical outcomes of the coronavirus disease 2019 (COVID-19) have remarkable potential for the breakdown of non-communicable diseases (NCDs) management by imposing ambivalence toward medication continuation. This study aimed to investigate the association of single or combinations of chronically used medications in NCDs with clinical outcomes of COVID-19.MethodsThis retrospective study was conducted on the intersection of two databases, the Iranian COVID-19 registry and Iran Health Insurance Organization. The primary outcome was death due to COVID-19 hospitalization, and secondary outcomes included length of hospital stay, Intensive Care Unit (ICU) admission, and ventilation therapy. The Anatomical Therapeutic Chemical (ATC) classification system was used for medication grouping. The frequent pattern growth algorithm was utilized to investigate the effect of medication combinations on COVID-19 outcomes.FindingsAspirin with chronic use in 10.8% of hospitalized COVID-19 patients was the most frequently used medication, followed by Atorvastatin (9.2%) and Losartan (8.0%). Adrenergics in combination with corticosteroids inhalants (ACIs) with an odds ratio (OR) of 0.79 (95% confidence interval: 0.68–0.92) were the most associated medications with less chance of ventilation therapy. Oxicams had the least OR of 0.80 (0.73–0.87) for COVID-19 death, followed by ACIs [0.85 (0.77–0.95)] and Biguanides [0.86 (0.82–0.91)].ConclusionThe chronic use of most frequently used medications for NCDs management was not associated with poor COVID-19 outcomes. Thus, when indicated, physicians need to discourage patients with NCDs from discontinuing their medications for fear of possible adverse effects on COVID-19 prognosis.

Summary: Recent advances in technology have resulted in generating or collecting massive volumes of data from rich data resources such as sensors and mobile devices in Internet of Things (IoT). Using data mining techniques can help overcome the mining problem in Fog computing environments which include millions of IoT devices. In addition, it can optimize response times, recourse consumption, and scalability in IoT applications. Frequent pattern mining, as one of the fundamental data mining tasks, is used for finding hidden patterns in such large datasets. The traditional data mining algorithms have many challenges such as scalability and resource consumption. This systematic review aimed to investigate the data mining algorithms, which focus on handling massive datasets, and present a technical taxonomy including the transaction‐centric, item‐centric, distributed, and parallel topics. The transaction‐centric and MapReduce‐based approaches were mostly utilized by 37% and 38%, respectively. Additionally, item‐centric, distributed, and parallel algorithms were employed 12% and 13%, respectively. The response time as a Quality of Service (QoS) factor had the highest percentage in the estimations of data mining algorithms (55%), followed by scalability (25%), and cost (20%). To the best of our knowledge, no study has focused on fog‐computing frequent pattern mining algorithms as one of the most important data mining tasks. This article aims to present a systematic review of the frequent pattern mining algorithms in fog computing and discuss the issues, challenges, and research perspectives for helping academia and industry leverage the power of data mining algorithms in fog computing. [ABSTRACT FROM AUTHOR]