CAP output report details

There are three types of sheet in every CAP report, Summary, Paragraph Segmentation and Result.

Comparative Assessment With Annotation on Paragraphs report details - Paragraph segmentation sheet

COLUMN NAME	DESCRIPTION
Paragraph #	Incremental paragraph index within the evaluated document.
Paragraph Id	Unique identifier of the paragraph ID per the original data import.
System 1 Engine	Name of the MT system 1.
System 2 Engine	Name of the MT system 2.
Domain	Domain classification of the evaluated content.
Additional information	Optional additional context or notes provided for the paragraph.
Metadata	Propagated metadata from input CSV (Optional).
Source paragraph	Original source paragraph text.
System 1 paragraph	Full paragraph translation generated by System 1.
System 2 paragraph	Full paragraph translation generated by System 2.
Segment #	Incremental segment index within the paragraph.
Source segment	Segmenter-defined segment created from the source paragraph, potentially spanning multiple chunks (e.g. sentences or similar linguistic units), and aligned with System 1 and System 2 segments. In case multiple chunks are present, they are separated by the ||| divider.
Source position	Position of the source segment within the source paragraph. The format is Start1-End1,Start2-End2, … For example when the segment starts at position 5 and ends on position 25, the field will be “5-25”.
System 1 segment	Segmenter-defined segment created from the System 1 translated paragraph, potentially spanning multiple chunks, and aligned with the source and System 2 segments. In case multiple chunks are present, they are separated by the ||| divider.
System 1 position	Position of the System 1 segment within the paragraph.
System 2 segment	Segmenter-defined segment created from the System 2 translated paragraph, potentially spanning multiple chunks, and aligned with the source and System 1 segments. In case multiple chunks are present, they are separated by the ||| divider.
System 2 position	Position of the System 2 segment within the paragraph. Same structure as for Source position
URL	Link to the particular Paragraph segmentation page.

Comparative Assessment With Annotation on Paragraphs report details - Result sheet

COLUMN NAME	DESCRIPTION
Paragraph #	Incremental paragraph index within the evaluated document.
Paragraph Id	Unique identifier of the paragraph ID per the original data import.
System 1 Engine	Name of the MT system 1.
System 2 Engine	Name of the MT system 2.
Domain	Domain classification of the evaluated content.
Additional information	Optional additional context or notes provided for the paragraph.
Metadata	Propagated metadata from input CSV (Optional).
Scale used	Scoring system ID.
Source paragraph	Original source paragraph text.
System 1 paragraph	Full paragraph translation generated by System 1.
System 2 paragraph	Full paragraph translation generated by System 2.
Segment #	Segment identification (generated incrementally).
Segment ID	Segment ID per the original data import.
Source segment	Source text for segment within paragraph.
System 1 segment	Translation from the system 1 for segment within paragraph.
System 1 Score	Score for Translation 1.
System 2 segment	Translation from the system 2 for segment within paragraph.
System 2 Score	Score for Translation 2.
Tie Breaker Result	Which system is better (values: 1, 2 or 0 in case both are of the same quality).
Judgement Duration (seconds)	Time spent on the segment judgement
Total Duration (seconds)	Time spent on the segment judgement.
URL	link to the particular segment evaluation.
Update Count	How many additional segment edits were done.
Last Updated	Timestamp of the last edit.
Last Updated By Evaluator	Name of the evaluator who made last edit.
Paragraph Id	paragraph identification(in case the contains the information is available)

In case the Job is configured with Annotation Set, the annotation values are added to the report. For each type of the text (source, translation1 or 2) the following details are added as new columns: each field header consist of combination of text type and annotation field.

For example, we have the Source annotation field named ‘Translation Issues’, This will be added to report as Source | Translation Issues column.