During learning, BIC uses your learnsets and the manually assigned candidates to identify the correlation between successful candidates, unwanted candidates and their environment. This correlation teaches BIC to automatically extract data that is similar to what was learned via the candidates.
Learning is required in the following cases.
- After you set up your extraction scheme and created learnsets
- When you added, deleted, moved, or renamed fields
- When you have changed any parameters affecting the analysis and evaluation methods
- When you have changed a learnset
Classification and extraction are usually learned together. To learn extraction, proceed as described for classification. For more information, see Use the Learn Mode.