Prediction Programming Concepts
The following topics describe the components of the Predictor resource.
Analysis Models
Two types of models are available. Prediction models are used for predicting user behavior, such as cross-sell patterns, and unknown user attributes. Segment models are used for classifying users into groups (segments) based on attributes and behavior, and for predicting the segment to which a particular user will most likely belong. For more information, see Analysis Models.
Analysis Model Configurations
The model configurations are contained in a set of database tables. These configurations contain information used to build analysis models, such as data sources and properties of the data atttributes. For more information, see Predictor Schema.
Prediction Model Viewer
Available through Commerce Server Manager in the Microsoft Managment Console (MMC), this browser provides a graphical representation of a prediction model. For more information, see Viewing Analysis Models.
PredictorClient Object
This object provides methods to predict user behavior, unknown user attributes, and the segments of a population to which a user belongs. For more information, see PredictorClient Object.
PredictorServiceAdmin Object
This object provides lists of available models and model configurations. For more information, see PredictorServiceAdmin Object.
PredictorServiceSiteAdmin Object
This object provides methods to rename and delete models and model configurations, and to generate a list of members of a segment. For more information, see PredictorServiceSiteAdmin Object.
PredModelBuilder Object
This object provides methods to control the building of models. For more information, see PredModelBuilder Object.
Segment Viewer
Available through Commerce Server Business Desk, this viewer provides a graphical representation of a segment model. Data for the Segment Viewer is provided by List Manager, which calls the GenerateSegmentList method of the PredictorServiceSiteAdmin. For more information, see Analyzing Population Segments.