Benchmarking is the process of evaluating the accuracy of a knowledge base model.
A catalog contains all the categories in the Siebel application that the knowledge base file that are used by Siebel Smart Answer to classify an incoming request.
A category is the type of a customer complaint or frequently asked question. Categories make up the Siebel catalog and are the input to the Siebel Smart Answer Server.
A collection of documents, emails, solutions to problems, manuals, or any other documentation that represents the business domain you want to capture and support in your Call Center production environment.
Any documents such as emails, solutions, manuals, or other documentation that are part of the Corpus is a Corpus Entry.
The hostname of a Siebel Smart Answer server concatenated with the knowledge base name: (for example, siebelserver99/KB)
Feedback is the positive re-enforcement of a correct categorization by the Siebel Smart Answer Server. This occurs explicitly when the user clicks Yes to "Was this useful" in the Contact Center, or when a category is selected in the auto-suggest mode of the Email Response Center.
Knowledge Base File
Knowledge base is the file that contains the set of categories that are part of the Siebel catalog.
Knowledge Base Model
Knowledge base model is the mapping of requests to categories.
A knowledge base is the combination of the knowledge base model and the knowledge base file.
Learning occurs when the Siebel Smart Answer engine incorporates feedback from the user into the concept model
Service Affecting Messages
Messages or requests that contain content indicative of a problem in the service or of overall dissatisfaction by the customer.
Training is the process of building a concept model for a category from a set of example emails or documents that represent the desired semantics of the category.