About Metrics
- Damian Rozmiarkowski (Unlicensed)
- Damian Rozmiarkowski
Introduction
XTM Metrics is one of the core components of every project created in XTM Cloud since it provides us with such vital information as:
project word count,
number of segments,
type and number of existing TM matches,
progress on individual workflow steps.
Additionally, since a single project can have multiple target languages and contain multiple files, XTM Metrics are further broken up into individual languages and files.
To learn more about the way in which word count is calculated in XTM Cloud, see How is wordcount calculated in XTM Cloud?.
Initial/In-between/Current Metrics
As the project workflow progresses, there is a reanalysis or continuous update, Metrics (word count, existing TM matches, etc.) change dynamically. With that in mind, it is necessary to distinguish between three types of Metrics:
Initial Metrics,
In-between Metrics (from other events in the project),
Current Metrics.
Initial Metrics
As the name suggests, Initial Metrics show Metrics that were generated at the time of project creation.
There is only one place in the XTM Cloud UI where you can find Initial Metrics. Go to Project editor → Metrics → Metrics history, and download the very first record called Metrics. You will receive an Excel sheet containing the data from the very beginning of a project.
In-between Metrics
In-between Metrics come from various events that might occur over the course of the project such as reanalysis or continuous project update.
They are stored in the same place as the Initial Metrics (Project editor → Metrics → Metrics history), and these are all the records except for the first one, Metrics, which is our Initial Metrics. Again, you will receive an Excel sheet containing data that presents the Metrics that applied at the time that a particular event took place in the project.
Current Metrics
As the name suggests, Current Metrics show Metrics which are up to date, i.e. they display data that is consistent with the current state of the project. This data changes dynamically as the project workflow progresses; for example, the progress bar gets longer with each new segment that is confirmed, and the TM count is updated when new TM matches are displayed in XTM Workbench.
Considering the above, what you see in the Metrics tab are Current Metrics and these are different from Initial Metrics, which is what one gets during the initial analysis.
In general, Current Metrics can be different if there is new information in the translation memory (TM) that did not exist during the mentioned analysis, hence it could not be included in initial metrics.
SAMPLE SCENARIO 1
Project Manager creates project A with 5 no-matching segments.
Project Manager creates project B with the same customer and the same 5 no-matching segments.
Currently both projects have the same metrics.
Project Manager assigns Linguists to both projects.
Linguist in project B translates all segments, they get added to the translation memory (TM).
Linguist opens XTM Workbench for project A in the Edit Mode. The segments get updated as there is new information in the translation memory (TM).
The Metrics now show 5 ICE segments.
The example above uses ICE segments, but the same would happen with Fuzzy Matches and Leveraged Matches, if there are similar segments.
The same might happen if you import the TM, add another customer’s TM to the project, and, in some situations, an additional TM match can come from the same project.
SAMPLE SCENARIO 2
Create a project with only file A.
Add file B to the project. This file should have the same exact source in some segments, but the context does not have to be the same.
Translate the same segment in file A.
File B will get an ICE/Leveraged Match.
This does not happen if you upload both files at the same time because file B would get cross-file repetitions. Of course, in this case, if Linguist starts with file B, then file A will have their Metrics changed, as an ICE/Leveraged Match would appear.
Current Metrics are always displayed in Project editor → Metrics.
Metrics vs. Statistics
It could be assumed that Metrics and Statistics are the same thing. In fact, they are two completely separate pieces of data.
Metrics
Metrics (Project editor → Metrics) provide information about the number of segments, words and characters in a project. Metrics also list the words and characters separately, as a percentage for the entire project, distributed into particular match types, matched against individual segments.
Metrics also provide the total translation progress, the number of segments, words and characters and the number of words and characters as a percentage, for "completed" segments per each workflow step, not taking specific linguists into account:
Statistics
Statistics (Project editor → Statistics), on the other hand, strictly relate to linguists employed in workflow steps and the amount of work they have done. They are generated when the linguist translates the segment or sets it as "Done". The match types displayed in Statistics reflect the segment's state at the moment of translation, in contrast to Metrics, which are dynamic and can change on the basis of new matches added to the TM. Statistics are divided into Source and Target statistics.
Source statistics → provide information about the number of segments, words, characters, and characters as a percentage in the source text, distributed into particular types of matches that were matched against individual segments, that have been translated in a specific workflow step by a specific user.
Target statistics → provide information about the number of segments, words, characters, and characters as a percentage in the target text, distributed into particular types of matches that were matched against individual segments, translated in a specific workflow step by a specific user.
Metrics-related reports
In XTM Cloud, a large number of Metrics-related reports are available for download.
Current Metrics for project
You can download a report that contains the Current Metrics for an entire project. This report is divided into tables, each for a separate language-file combination.
Go to Project editor → Metrics → Save all as, and choose the format in which to download the report: XLSX, Flat XLSX or CSV.
Current Metrics for individual file
You can download a report that contains the Current Metrics for a particular target language, for an individual file. This report is based on a table whose data you are downloading.
Go to Project editor → Metrics, select a relevant target language and find the Metrics for a file you are interested in. At the bottom, you can choose one of two file formats: XLSX or CSV.
The same report can also be downloaded directly from XTM Workbench (File → Metrics → Download).
Metrics history
As the name suggests, this type of report displays all the Metrics events that were generated during a specific project. Each Metrics generation event is presented in a separate Excel worksheet. You can download the Metrics history for multiple projects at once. If you do so, you download a ZIP package containing several Excel files, each for one project.
Go to Projects → Project list → Actions → (select Metrics history from a dropdown) → (include relevant projects), and click Run.
You can also specify the particular time frame in which the relevant Metrics events were created; either the whole duration of the project or a specific time period.
Combined metrics
As the name suggests, a Combined metrics report aggregates the Current Metrics from all projects selected at the time the report was generated. The output file is a single Excel worksheet containing a table and, at the top, the names of projects that were included in the report.
Go to Projects → Project list → Actions → (select Combined metrics from a dropdown) → (include relevant projects), and click Run.
Costs and Metrics reports
There are also reports that contain some Metrics-related data but are more oriented towards costs relating to particular linguists. To learn more about them, we recommend that you read this article: Costs & Metrics reports.