2
$\begingroup$

I am new to NLP, I found a format named ConLL which seems a tab-separated file, like

ID FORM LEMMA PLEMMA POS PPOS FEAT PFEAT HEAD PHEAD DEPREL PDEPREL 

I found a dependency Treebank in this format for Persian. I would like to do some experiment on it. I am familiar with C# and C# has great libraries for XML document files.

I just wonder why such a corpus is not in XML format! I don't say a sophisticated one but they could put each line in a node of XML. Then at least, I would know what is the tag for each element.

However, I can convert it to XML.

$\endgroup$
1
$\begingroup$

One reason for sure is, that you can easialy open it in spreadsheet viewer.

$\endgroup$
1
$\begingroup$

ConLL is a format optimised for processing efficiency (both speed and memory usage).

XML formats for NLP often have the annotation layers one after the other; therefore you have to hold the whole file in memory for processing it.

A drawback of ConLL is that you have to put in fillers (e.g., 0) for sparse annotations.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.