I am ingesting via csv crawl and have three questions:
do i need to convert first in the csv the dates into an epoch format to be able to use the filter by date?
do i really have to use group_by? it’s a bit confusing to me what should i choose as a ‘group’ and if i go by individual name (what i want to see in the title of the snippet), the upload is getting much much longer.
do i understand correctly that i can choose to show both the snippets and the full text where the snippet is from?
Thank you
do i need to convert first in the csv the dates into an epoch format to be able to use the filter by date?
For the time being, yes. We do plan to introduce a formal date format later.
do i really have to use group_by? it’s a bit confusing to me what should i choose as a ‘group’ and if i go by individual name (what i want to see in the title of the snippet), the upload is getting much much longer.
I’m not sure exactly what you’re referring to. Can you give a bit more detail on where you’re seeing group_by?
do i understand correctly that i can choose to show both the snippets and the full text where the snippet is from?
There are 2 things you can do:
If you want snippets + surrounding context, have a look at the highlighting docs here: Highlighting and Snippet Extraction | Vectara Docs . You can provide a number of sentences or characters that you’d like to include before/after the relevant snippet. There’s also a complete example with code samples at Query | Vectara Docs if you’d like to try
If you “just” want to know where the document came from, we’d recommend using the document ID for that, which is returned with each result