Problema con Get OCR Text

I am validating an exercise that allows me to read some exams in PDF format. From this document I select the results of the tests. And from there I compare them with the exam results found in a table in Excel. The problem is with the OCR recognition method, since every time I execute it, certain information varies, it is not very accurate.

I would like to know what alternatives I have so that the information I am extracting is more reliable and truthful.
Main.xaml (56,4 KB)

Fuentes de informacion
Book2.xlsx (17,2 KB)
Evaluation.pdf (383,5 KB)