Where is the Doc Extractor?

Help Me Build
1

I am trying to build a multimodal RAG using nova-2-multimodal-embeddings-v1:0.
When I processed a PDF file containing both images and text with Dify Extractor, I was only able to obtain textual information.

Based on the following documentation, I believe that I need to use Doc Extractor instead of Dify Extractor:
https://docs.dify.ai/en/use-dify/knowledge/knowledge-pipeline/knowledge-pipeline-orchestration#doc-extractor

However, I cannot find Doc Extractor in my environment at https://cloud.dify.ai/.
Where is this Data Processing Tool located?

image
image1920×1013 224 KB

2

image
image933×764 53.4 KB

2 Likes
3

Thank you. I was able to find it by changing the language setting to English.

However, even when using this, it seems that only the text contained in the PDF can be extracted.
I don’t know how to turn a document that contains a mix of images and text into a knowledge base while preserving its meaning.

With Amazon Bedrock Knowledge Bases, you can specify a “foundation model as a parser” in the “parsing strategy,” but…

4

image
image1246×768 68.7 KB

According to the documentation, the Dify extractor plugin is what you need. I haven’t used it yet, but I’ll give it a try later.