What is Java Data Mining, JDM?

Question

I am looking at JDM. Is this simply an API to interact with other tools that do the actual data mining? Or is this a set of packages that contain the actual data mining algorith

Answer 1

This project is inactive. It is an API. It does not provide algorithm implementations.

Answer 2

AFAIK, this is just an API to link Java applications with data mining tools. See the actual API and this whitepaper for more details.

Here's the package list from the JDM API (can't seem to find it online):

• javax.datamining This package contains Java classes and interfaces describing basic objects used in the other JDM packages.
• javax.datamining.algorithm.feedforwardneuralnet This package contains definitions for describing the algorithm settings specific to building a fully connected, n-layer, feed-forward neural network model.
• javax.datamining.algorithm.kmeans This package contains definitions for describing the algorithm settings specific to building a k-means clustering model.
• javax.datamining.algorithm.naivebayes This package contains definitions for describing the algorithm settings specific to building a Naive Bayes model.
• javax.datamining.algorithm.svm This package is Algorithm.SVM.
• javax.datamining.algorithm.svm.classification This package is Algorithm.SVM.Classification.
• javax.datamining.algorithm.svm.regression This package is Algorithm.SVM.Regression.
• javax.datamining.algorithm.tree This package contains definitions for describing the algorithm settings specific to building a decision tree model.
• javax.datamining.association This package contains Java classes describing the settings and model for the association mining function.
• javax.datamining.attributeimportance This package contains Java classes describing the settings and model for the attribute importance mining function.
• javax.datamining.base This package contains Java interfaces that are the top-level objects in JDM and consist of named objects as well as top-level base interfaces such as ModelDetail and AlgorithmSettings.
• javax.datamining.clustering This package contains Java classes describing the settings, model and apply output for clustering.
• javax.datamining.data This package contains Java classes describing mining data including the physical and logical aspects of data as well as the definition of matrix.
• javax.datamining.modeldetail.feedforwardneuralnet This package contains Java classes describing the algorithm-related details of a feed forward neural network model.
• javax.datamining.modeldetail.naivebayes This package contains Java classes describing the algorithm-related details of a Naive Bayes model.
• javax.datamining.modeldetail.svm This package is ModelDetail.SVM.
• javax.datamining.modeldetail.tree This package contains Java classes describing the algorithm-related details of a tree model.
• javax.datamining.resource This package contains Java classes used to create the connection to the data mining engine (DME) and maintain metadata objects associated with the DME.
• javax.datamining.rule This package contains Java interfaces representing the predicate rules created as part of the models such as tree model.
• javax.datamining.statistics This package contains Java classes describing the statistics of the input mining data.
• javax.datamining.supervised This package contains Java classes describing the generic settings, model, test task and test result of supervised learning.
• javax.datamining.supervised.classification This package contains Java classes describing the settings, model, and test task and result for classification.
• javax.datamining.supervised.regression This package contains Java classes describing the settings, model, and test task and result for regression mining function.
• javax.datamining.task This package contains Java classes that define build mining task, a primary mining operation that builds mining models.
• javax.datamining.task.apply This package contains Java classes that define the apply mining task for applying a mining model.

Answer 3

Wikipedia says:

Java Data Mining (JDM) is a standard Java API for developing data mining applications and tools. JDM defines an object model and Java API for data mining objects and processes.

According to this article and the JSR for JDM 2.0 (#247):

By extending the existing JDM standard with new mining functions and algorithms, data mining clients can be coded against a single API that is independent of the underlying data mining system. The goal of JDM is to provide for data mining systems what JDBCTM did for relational databases.

So it appears that, yes, JDM is an API to interact with other tools that do the actual mining. It also appears that this JSR is currently inactive.

Answer 4

Ah, the wonders of the interweb:

Java Data Mining (JDM) is a standard Java API for developing data mining applications and tools. JDM defines an object model and Java API for data mining objects and processes. JDM enables applications to integrate data mining technology for developing predictive analytics applications and tools. The JDM 1.0 standard was developed under the Java Community Process as JSR 73. As of 2006, the JDM 2.0 specification is being developed under JSR 247.

Lists some implementations also, although it looks like it may be a dead duck.