html content leads to errors while the same content in text format is processed fine
Posted on: Wed, 06/08/2011 - 05:13
Hi,
I need to get the topic end entities information for an ad-hoc user-supplied string in an application. So I tested it with the OpenCalais desktop submission tool. An example I tried was 'Japan earthquake'. Now because OpenCalais requires the content to be at least 100 characters long so I repeat the same text multiple times to make the string long enough. I get the semantic metadata information. However, if I submit the same text as html:something like-
<html>
<body>
Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake Japan earthquake
</body>
</html>
I get a response:
<Exception>Calais continues to expand its list of supported languages, but does not yet support your submitted content.</Exception>
Thanks
ydn_dev
Trackback URL for this post:
http://www.opencalais.com/trackback/106642
Does anyone know how to ask the SemanticProxy URL string to use the french or spanish entity extraction protocol? It defaults to english, but I would like to test the other two languages as well. I would appreciate any guidance on this. Regards,
It works on French and Spanish langugaes as well. You just have to sent the content as you would have done for English content.
You can use Content-Type: text/htmlraw instead of text/html and you won't get the exception.
General note - you can send to open calais text which contains less than 100 characters, in those cases you might get less result on your output but you will not get an exception.