CSV knowledge not working for me

[Igorgro]

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• Please do not modify this template :) and fill in all the required fields.

1. Is this request related to a challenge you're experiencing? Tell me about your story.

I'm trying to make a simple knowledge from the CSV file I've made. This is the file for reference:
cat_breeds.csv. This is a slightly reformatted table from this Wikipedia article. When I upload this file to knowledge, Dify correctly splits it into chunks:

I'm not sure if this the best way to split this file into chunks, but changing CSV settings (semicolon instead of comma, etc) makes the chunking get completely broken. After vectorizing and saving the file into knowledge, I'm trying to make a simple test, but getting totally irrelevant results:

I want to know, if this a problem with my data, setup, or I don't fully understand how RAG and vector search should work?

2. Additional context or comments

This is my setup:

1. Dify in Docker
2. Database: PostgreSQL 17
3. pgvector vector storage on the same PostgreSQL server
4. Embedding LLM: Ollama with nomic-embed-text:latest model
5. Knowledge settings: