Support for users of CCG software closed 7-27-20

[Lev-Arie Ratinov <why2ask AT gmail.com>]

Hi Zhang. Thanks for the interest.

The NER package will require a lot of memory for a long text. However
for reasonably-sized texts (e.g. news article, Wikipedia page), 1.7G
should be plenty.

I've graduated 2 years ago, and no longer support the system.
Moreover, I'm not aware of any "runDemo.sh" script. It may have been
added since I've graduated. I've never ran my system on an EC2 server,
but I recommend you to run it locally first, and see why/who it fails.
If you add a more informative error message, there's more chance
someone will help you.

illinois-ml-nlp-users (mailing list attached) is a good mailing list
and discussion forum which will allow you to solve your problems.

Best.
Peace&Love


On Sun, Jun 16, 2013 at 10:51 PM, Zhang, Zhu - (zhuzhang)
<zhuzhang AT email.arizona.edu>
 wrote:
> I figure I'd better use your gmail address :)
>
>
>
> ________________________________
> From: Zhang, Zhu - (zhuzhang)
> Sent: Sunday, June 16, 2013 7:42 PM
> To: 
> ratinov2 AT uiuc.edu
> Subject: Question about Illinois Named Entity Tagger
>
> Dear Lev,
>
>
>
> The iNER seems nicely built based on your publications. However, I had
> troube running it and producing results. Two questions:
>
>
>
> 1. Does it require a lot of memory? Running the "runDemo.sh" on an EC2
> server with 1.7G memory wasn't successful at all, and the final output was
> the following:
>
>
>
> Tagging file: SampleInputs/longParagraph.txt
> Reading model file : ./data/ner-ext/Models/Ontonotes/Ontonotes.model.level1
> Killed
>
> 2. Given the resource consumption and initialization time, can the program
> run as some sort of daemon? If not, what are your recommendations on
> processing large number of text files?
>
>
>
> Thanks in advance for your help!
>
>
>
> Drew