2 A Tour of the Data Mining APIs

Build Results

The CREATE_MODEL procedure creates a mining model. The attributes, transformations, rules, and other information internal to the model are returned by GET_MODEL_DETAILS functions. You can also obtain information about mining models by querying data dictionary views, as described in "Data Mining Data Dictionary Views".

Apply Results

The APPLY procedure creates a table with specific columns and populates the columns with mining results. The columns of this table vary based on the particular mining function and algorithm.

Physical Data Set

Physical data sets refer to the data to be used as input to data mining operations. Physical data set objects reference specific data via a URI, e.g., specifying a table or file. Oracle supports a table or view in the same database as a valid physical dataset URI. Syntax of the oracle physical dataset URI is as follows:

Data URI Syntax:

[schemaName.] tableName/viewName

The physical data set object can support multiple data representations. Oracle Data Mining supports two types of data representation, i.e., single-record case and wide data. Refer to data mining concepts guide for more details. The Oracle implementation requires users to specify the case-id column in the physical dataset.

A physical data set object is a transient object in the Oracle Data Mining Java API. It is stored in the Connection object as an in-memory object.

Build Settings

A build settings object captures the high-level specification input for building a model. The API specifies mining functions: classification, regression, attribute importance, association, clustering, and feature extraction.

Build settings allows a user to specify the type of result desired without having to specify a particular algorithm. Although a build settings object allows for the specification of an algorithm and its settings, if the algorithm settings are omitted, the DME selects an algorithm based on the build settings and possibly characteristics of the data.

Build settings may also be validated for correct parameters using the verify method.

Build settings is a persistent object in the API, it is stored as a table with the user specified name in the user schema. This settings table is interoperable with the PL/SQL API. It is recommended not to modify the build settings table manually in the schema.

Task

The execute method in the Connection object is used to start an execution of a mining task. Typically, mining operations are done using tables with millions of records, so the execution of operations like building a model can take more time. JDM supports asynchronous execution of mining tasks using DBMS_SCHEDULER in the database. Each mining task is stored as a DBMS_SCHEDULER Job object in the user schema. When the user saves the task object, it creates a Job object and sets the object to be in the DISABLED state. When the user executes a task, it enables the job to start execution.

Applications that would like to use the DBMS_SCHEDULER functionality like scheduling mining tasks could use DBMS_SCHEDULER PL/SQL package provided in the database.

To monitor tasks that are executed asynchronously, the execute method returns a javax.datamining.ExecutionHandle object. It provides all the necessary features like waitForCompletion, getStatus methods to retrieve the status details of the task.

Model

A model object is the result of applying an algorithm to data as specified in a build settings object.

Models can be used in several operations. They can be:

• Inspected, for example to examine the rules produced from a decision tree or association

• Tested for accuracy

• Applied to data for scoring

• Exported to an external representation such as native format or PMML

• Imported for use in the DMS

When a model is applied to data, it is submitted to the DMS for interpretation. A Model references its BuildSettings object as well as the Task that created it. as.

Test Metrics

A Test Metrics object is the result of testing a supervised model with test data. Based on the type of mining function, different test metrics are computed. For classification models, accuracy, confusion-matrix, lift, and receiver-operating characteristic can be computed to access the model. Similarly for regression models, R-squared and RMS errors can be computed.

Apply Settings

An apply settings object allows users to tailor the results of an apply task. It contains a set of ordered items. Output can consist of:

• Data to be passed through to the output from the input dataset, e.g., key attributes

• Values computed from the apply itself, e.g., score, probability and in the case of decision trees, rule identifiers

• Multi-class categories for its associated probabilities, e.g., in a classification model with target favoriteColor, users could select the specific colors to receive the probability that a given color is favorite.

Each mining function class defines a method to construct a default apply settings object. This simplifies the programmer's effort if only standard output is desired. For example, typical output for classifications apply would include the top prediction and its probability.

Transformation Sequence

A transformation sequence object represents the sequence of transformations that are to be performed as part of a mining operation. For example, a Support Vector Machine model build involves outlier handling and normalization transformations. In addition to this, there can be new derived attribute creation and business transformations, and so on. Typicaly these transformations are resused for other model builds and hence applications can save the transformation sequence as a named object in the API.

Transformation sequence can be used either to perform transformations as a separate task or embed them to the modeling process by specifying it as one of the input objects for model building.