If the new line is present it will be ignored by the name finder.
Adding a special token for it can help if your data has a certain
format where the new line char is important to detect the names (e.g.
lists where names are written in each line).

I suggest you try it and see how it changes the performance on your data.


On Thu, Apr 12, 2018 at 12:09 PM, Damiano Porta <damianopo...@gmail.com> wrote:
> Pardon, i did not explain "where"... i am talking about the training of a
> NER model
> 2018-04-12 12:05 GMT+02:00 Damiano Porta <damianopo...@gmail.com>:
>> Hello,
>> i need new lines in my document. Should i escape it with a custom token
>> like <n> ?
>> Thanks

Reply via email to