|
Log
|
A structured record logging a single event in a system with no rich attribute.
Examples:
- an HTTP server log line,
- a single SAP transaction,
- a single line of a purchase receipt,
- an email (excluding attachments).
|
1 token = 1000 logs
|
|
External Document
|
An unstructured document crawled from websites, which do not belong to the
customer.
Examples:
- an HTML web page,
- a PDF document crawled from a website that does not belong to the customer.
|
8 tokens = 1000 external documents
|
|
Internal Document
|
An unstructured document created by the customer.
Examples:
- an internal wiki page,
- a page of an internal knowledge base portal (even if retrieved using the web
crawler),
- bulk office documents and files on employee hard disk drives.
|
14 tokens = 1000 internal documents
|
|
Business Item
|
A business document, work product or commodity, used in business operations by the
customer.
Examples:
- an article in ERP with quantities, price, sourcing;
- a customer record in CRM with inventory, contracts;
- one ERP item in a CSV file grouping a set of ERP items;
- a structured office document, for example, a reference document or final
version of a document;
- a manufactured product processed or sold by the company: part, assembly with
requirements, material, weight etc.
|
60 tokens = 1000 business items
|