Effectively and Efficiently Mining Frequent Patterns from Dense Graph Streams on Disk

Braun, Peter; Cameron, Juan J.; Cuzzocrea, Alfredo; Jiang, Fan; Leung, Carson K.

doi:10.1016/j.procs.2014.08.114

Cited by 45 publications

(24 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It would be interesting to see how much the performance increase if other features are used, hence leading to a feature hyperspace as indicated in Section 3. Another possible direction of research consists in making our algorithm compliant with emerging features of novel big data systems (e.g., [18,28,3,29]).…”

Section: Final Remarks and Future Workmentioning

confidence: 99%

Event Detection in Complex Environments: An Effective and Efficient Machine-Learning-Based Framework (S)

Cuzzocrea¹,

Mumolo²

2019

International Conferences on Distributed Multimedia Systems

Self Cite

View full text Add to dashboard Cite

In this paper we describe a falls detection and classification algorithm for discriminating falls from daily life activities using a MEMS accelerometer. The algorithm is based on a shallow Neural Network with three hidden layers, used as fall/non fally classifier, trained with daily life activities features and fall features. The novelty of this algorithm is that synthetic falls are generated as multivariate random Gaussian features, so only real daily life features must be collected during some day of normal living. Moreover, the features related to synthetic fall events are generated as complement of normal features. First of all, the features acquired during daily life are clustered by Principal Component Analysis and no Fall activities shall be recorded. The complement set of the normal features is found and used as a mask for Monte Carlo generation of synthetic fall. The two feature sets, namely the features recorded from daily life activities and those artificially generated are used to train the Neural Network. This approach is suitable for a practical utilization of a Neural Network based fall detection characterized by high Recall-Precision rate.

show abstract

Section: Final Remarks and Future Workmentioning

confidence: 99%

Event Detection in Complex Environments: An Effective and Efficient Machine-Learning-Based Framework (S)

Cuzzocrea¹,

Mumolo²

2019

International Conferences on Distributed Multimedia Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…An efficient stream mining algorithms was designed in [17] to mine frequent patterns effectively and efficiently from such streaming data.In [18], a tree-based mining algorithm was developed that used to mine frequent patterns from dynamic uncertain data stream based on time-fading and landmark models.An efficient Skyline query algorithm was designed in [19] of uncertain moving stream data but the performance is failed to use index structures such as grid, R tree and so on. A novel algorithm was effectively designed in [20] to mine the uncertain group patterns that improve the mining efficiency but it not capable to handle more complex patterns.…”

Section: Related Workmentioning

confidence: 99%

Fusion Layer Topological Space Query Indexing For Uncertain Data Mining

Kalavathi¹,

Suresh²

2016

IJECS

View full text Add to dashboard Cite

-Data uncertainty is an intrinsic property in different applications such as sensor network monitoring, object recognition, location-based services (LBS), and moving object tracking. The data mining methods are applied to the above mentionedapplications their uncertainty has to be handled to achieve the accurate query results. The several probabilistic algorithm estimates the location and control for each object but not effective in handling query processing in distributed environments. Probabilistic inverted indexing computes the lower bound and upper bound for a threshold keyword but fails to extend the technique on tackle the correlation. In order to overcome the issues in uncertain data mining, Fusion Layer Topological Space Query Indexing technique (FLTS)is introduced. Initially, the queries are articulated on any random subset of attributes in the uncertain data.The FLTS index technique answers the top-k queries competently. FLTS index correctlyshows in their dominant relationships andsignificantly reduces the number of tuple accessed through query processing by pruning redundant tuple based on two criterions such as layer point sort method and the record point sort method. Initially, layer point sort method is used in FLTS indexto sort out tuple totally on the combination of all attribute values of the tuple. Subsequently, each attributes values particularly used for rating the tuple using record point sort method. Therefore,the correlation is removed significantly. Through an analysis of the interaction of the two sorting methods, derive a fixed bound that reduces the number of tuple retrieved during query processing and obtaining the correct query results in distributed environments. An experimental result shows that the FLTS indexing technique improves the query retrieving efficiency, response time, memory consumption and scalability.

show abstract

“…Various approaches, such as graph clustering, graph stream classification, subgraph mining, and frequent subgraph pattern detection, have been proposed for graph analysis [ 5 , 21 , 22 , 23 , 24 , 25 , 26 , 27 ]. Frequent subgraph pattern, which detects a subgraph frequently occurring during a specific period is a widely used analysis method for graph streams [ 28 , 29 , 30 , 31 , 32 ]. In the IoT environment, frequent subgraph pattern is used for analyzing interactions among various objects or for determining anomalies [ 33 , 34 , 35 ].…”

Section: Introductionmentioning

confidence: 99%

“…Once the graph stream data are inputted, a DSTree is constructed, using whichever FP-tree was constructed to detect the frequent pattern. In Reference [ 31 ], Data Stream Matrix (DSMatrix) was proposed for storing graphs more efficiently than the DSTree proposed in Reference [ 30 ]. The DSMatrix is a two-dimensional array, so it can be constructed at a lower cost than that of DSTree.…”

Section: Introductionmentioning

confidence: 99%

Detecting Incremental Frequent Subgraph Patterns in IoT Environments

Bok

Jeong

Choi

et al. 2018

Sensors

View full text Add to dashboard Cite

As graph stream data are continuously generated in Internet of Things (IoT) environments, many studies on the detection and analysis of changes in graphs have been conducted. In this paper, we propose a method that incrementally detects frequent subgraph patterns by using frequent subgraph pattern information generated in previous sliding window. To reduce the computation cost for subgraph patterns that occur consecutively in a graph stream, the proposed method determines whether subgraph patterns occur within a sliding window. In addition, subgraph patterns that are more meaningful can be detected by recognizing only the patterns that are connected to each other via edges as one pattern. In order to prove the superiority of the proposed method, various performance evaluations were conducted.

show abstract

Effectively and Efficiently Mining Frequent Patterns from Dense Graph Streams on Disk

Cited by 45 publications

References 27 publications

Event Detection in Complex Environments: An Effective and Efficient Machine-Learning-Based Framework (S)

Event Detection in Complex Environments: An Effective and Efficient Machine-Learning-Based Framework (S)

Fusion Layer Topological Space Query Indexing For Uncertain Data Mining

Detecting Incremental Frequent Subgraph Patterns in IoT Environments

Contact Info

Product

Resources

About