In a saved search using a training data source, what is typically the only field returned as a column?

In a saved search using a training data source, the field that is typically returned as a column is the Extract Text. This is primarily because Extract Text represents the content extracted from the documents that have been analyzed during the training phase. The Extract Text field allows users to view the actual text extracted from the source materials, which is crucial for reviewing and assessing the information present within the documents.

This focus on Extract Text serves functional purposes within the analytics process, especially in contexts where understanding the content is key to training models or refining search parameters. Summarizing data in this way ensures that users can evaluate and use the included text for further training and assessment of the models, making it an integral part of the analysis.

While Document ID, Document Title, and Keywords provide useful information, their roles are generally more about identification and categorization rather than conveying substantive content. Document ID is for unique identification, Document Title gives a name to the document, and Keywords serve more as metadata that assists in the search process but do not present the detailed content necessary for analysis in the same way that Extract Text does.