GeoDenStream

GeoDenStream is a method for clustering spatiotemporal data streams. Building on DenStream, this method is particularly suitable for analyzing entity-based geographical data streams.

spatiotemporaldata streamsDenStreamentity-basedgeographical data streams

true

Contributor(s)

Initial contribute: 2020-01-09

Authorship

:  
View
Is authorship not correct? Feed back

Classification(s)

Method-focused categoriesData-perspectiveGeoinformation analysis
Method-focused categoriesData-perspectiveGeostatistical analysis

Detailed Description

English {{currentDetailLanguage}} English

Quoted from: https://github.com/manqili/GeoDenStream 

GeoDenStream is an improved DenStream clustering method for acquiring individual data point information within Big Data Streams. The implementation of GeoDenStream is based on the open source MOA project (Bifet et al., 2010), which is available at (https://github.com/Waikato/moa).

In GeoDenStream, several modifications and improvements are made based on the MOA package to address the memory limitation, overlapping points, and false noise issues. Also, pruning strategy is extended to consider the time stamp of the data records.

Detailed modifications and improvements can be found in master/moa/src/main/java/moa/clusterers/denstream. There are four main classes in this folder:

  • DenPoint.java This class is same with the original MOA, which is used to record a point(aka a record, an item, a tweet) in a stream.
  • MicroCluster.java In this class, a new function TryInsert was add for attempting add a point into a potential cluster and check whether it can be added. Two lists are added for recording: (1) Ids of the points that belongs to a potential cluster and (2) the distance between each involved point and the center of a potential cluster (could be seen in Line 36 and 37).
  • TimeStamp.java This class is same with the original MOA, and it is used for giving a timestamp as long integer for each points.
  • WithDBSCAN.java The code in this class has been modified largely. Function trainOnInstanceImpl is a new implemented function which supports the real time based pruning strategy. And function trainOnInstanceImpl_TpStaticIndex keeps the count based pruning strategy as MOA does. A minor issue but also a key point. the function nearestCluster was modified. The original nearestCluster has a bug in calculating the minimum distance. Could be seen in Line 483.

Examples of its application using Twitter data streams can be found in master/moa/src/main/java/denstream/zikaebola.


For other analysis scenarios, a configurable application is provided in master/moa/src/main/java/denstream/configure.

  • DSC_Dynamic.java This is an entrance for using GeoDenStream, with dynamic memory optimization (loading points with an offline range and using the index strategy).
  • DSC_Static.java This is another entrance for using GeoDenStream, without memory optimization (loading all records for generating clusters).
  • ProcessPotentialCluster.java This class is used for handling the overlap and false noise issues. Detailed implementation could be checked in the funciton fillPotentialCluster.
  • ProcessOfflineCluster.java This class is for get offline clusters based on the results of ProcessPotentialCluster.java. For other classes, they are some utilies and could be easy understand.

{{htmlJSON.HowtoCite}}

ManqiLi, ArieCroitoru, and SongshanYue (2020). GeoDenStream, Model Item, OpenGMS, https://geomodeling.njnu.edu.cn/modelItem/b9068a36-81f6-4748-815c-b31311e99130
{{htmlJSON.Copy}}

Contributor(s)

Initial contribute : 2020-01-09

{{htmlJSON.CoContributor}}

Authorship

:  
View
Is authorship not correct? Feed back

History

Last modify time
2021-01-09

QR Code

×

{{curRelation.overview}}
{{curRelation.author.join('; ')}}
{{curRelation.journal}}









{{htmlJSON.RelatedItems}}

{{htmlJSON.LinkResourceFromRepositoryOrCreate}}{{htmlJSON.create}}.

Drop the file here, orclick to upload.
Select From My Space
+ add

{{htmlJSON.authorshipSubmitted}}

Cancel Submit
{{htmlJSON.Cancel}} {{htmlJSON.Submit}}
{{htmlJSON.Localizations}} + {{htmlJSON.Add}}
{{ item.label }} {{ item.value }}
{{htmlJSON.ModelName}}:
{{htmlJSON.Cancel}} {{htmlJSON.Submit}}
Name:
Version:
Model Type:
Model Domain:
Scale:
Purpose:
Principles:
Incorporated models:

Model part of

larger framework

Process:
Information:
Initialization:
Hardware Requirements:
Software Requirements:
Inputs:
Outputs:
{{htmlJSON.Cancel}} {{htmlJSON.Submit}}
Title Author Date Journal Volume(Issue) Pages Links Doi Operation
{{htmlJSON.Cancel}} {{htmlJSON.Submit}}
{{htmlJSON.Add}} {{htmlJSON.Cancel}}

{{articleUploading.title}}

Authors:  {{articleUploading.authors[0]}}, {{articleUploading.authors[1]}}, {{articleUploading.authors[2]}}, et al.

Journal:   {{articleUploading.journal}}

Date:   {{articleUploading.date}}

Page range:   {{articleUploading.pageRange}}

Link:   {{articleUploading.link}}

DOI:   {{articleUploading.doi}}

Yes, this is it Cancel

The article {{articleUploading.title}} has been uploaded yet.

OK
{{htmlJSON.Cancel}} {{htmlJSON.Confirm}}