Any way to exclude errors for arabic numbers and spelled out numbers


#1

Hello.
I would like to ask for your help in the situation below.

<Example 1>
Source: double
Target: 2

<Example 2>
Source: 22
Target: Twenty-two

Although these are not an error, simillar texts keep coming up as an error.
Since there are so many of them, I end up ovelooking the actual numeric errors.

Is there any way to avoid this?


#2

Hi Maru,

These false alarms cannot be prevented via Numeric Mismatch QA check. However, you can check these type of errors using checklists.

Example 1:
Source: “double” OR “two”
Target: -“2”
Powersearch: ON
Search mode: simple
Match Whole Word checkbox: enabled

Example 1:
Source: “22”
Target: -“twenty-two”
Powersearch: ON
Search mode: simple
Match Whole Word checkbox: enabled

These searches will display any segment that contains the Arabic number but is missing the spelled out number.

For more information about checklists, I would recommend you to watch these two videos from the Xbench youtube channel:

  1. Using ApSIC Xbench 3.0 Checklists for QA (Part 1)
  2. Using ApSIC Xbench 3.0 Checklists for QA (Part 2)

Regards,
Oscar.


#3

Thank you for your fast reply!
I see…then I’ll try making a checklist for these errors.
Thank you for the video.