strange ?question?mark characters
strange ?question?mark characters
Posted on: Tue, 06/30/2009 - 12:42
I am getting a lot of nodes with returned to me with labels in this format "?the?designers?of?this?fun?product"
I have a drupal blog setup with UTF8, and this is happening only since the new 4.1 API release. Wondering how to solve it. Did anyone else see something similar?
Trackback URL for this post:
http://www.opencalais.com/trackback/24762

Hi Anand,
The strange organization and characters issue should have been fixed. Can you please verify.
sumit
Hi Anand,
Are you using Tagaroo with your Drupal blog? I am asking because the question is posted in the SemanticProxy forum, so wanted to confirm.
We have not heard any such issues. Can you share the sample text which is returning these strange lables?
sumit
hi sumit, i am using opencalais on drupal, with semantic proxy enabled.
here are a couple of error sources:
http://www.affr.nl/news/196/Fountain_head_revisited/
http://archiblog.d-earle.com/2009/06/17/furniture/
thanks
Hi Anand,
Thanks for pointing out the issue. Some of the internationalization libraries on one of our servers were corrupted. We have fixed this and everything should be fine now. Can you please check and confirm?
sumit
Hi Sumit, this is perhaps still broken, the "?" marks are gone but I get some very strange phrases, for example when I resaved http://www.spaceandculture.org/2009/06/22/you-are-the-city/ I got Organization: "and transformation of urban settings" at Calais threshold of .33.
Here is a more illustrative one,
URL: http://fantasticjournal.blogspot.com/2009/06/lines-of-defence.html
has an Organisation as follows: "In a field near my parent's house in rural Essex there is a circle of trees. In the centre of this circle is a deep pit which for as long as I can remember has been an overgrown dump full of old pesticide canisters"
Also, a couple of names on the page are not classed under people, as they used to - those just disappeared.
Hi Anand,
The issue we had with the '?' characters is fixed for sure, this is why we do not see strange characters in the extraction any more.
The wrong organization seems to be an different issue, what we refer to as 'extraction' issue. I will let the team know about this and they will shoot for a fix asap.
Thanks.
Sumit
Hi Sumit, I am using Opencalais plugin for Drupal with Semantic proxy switched on. The errors only appear when I choose the use semantic proxy option.