Solved

Arabic PDF extract learning instance in Community IQ bot

  • 3 January 2023
  • 2 replies
  • 71 views
R
Badge +2

Hi Team , 

I have requirement to extract text from Arabic invoice pdf documents and wanted to try with IQ bot learning instance for the feasibility . Unfortunately ,  Arabic language isn`t visible in the primary language selection while creating instance in the IQ bot . 

 

Could you help with the other ways to try the Arabic pdf extract using IQ bot. 

 

Thanks in advance 

 

icon

Best answer by Padmakumar 3 January 2023, 07:20

View original

2 replies

Padmakumar
Userlevel 7
Badge +13

Hi @Rajeswari.N1 ,

 

IQ Bot does support Arabic but unfortunately, it falls in the document category of Others. Means, you can’t extract Arabic text if your document type is invoice, contracts, health insurance, purchase order, and so on. 

 

Please refer here for further details on this.

 

But you can try changing the OCR engines for the same. For Microsoft Azure Computer Vision OCR engine, user can select any language from IQ Bot's drop-down, but the API aims to auto-detect the language during processing, and override user selection.

 

I hope this will help.

Padmakumar
P
Badge

Hi @Rajeswari.N1 - indeed, mentioned documents types are not supported, but you can try other OCRs, to share personal experience - we are using Sanad.ai for structured data extraction like invoices.

Reply


Terms & ConditionsCookie settings