[sw-l] Broadcast corpus

[Steve Slevinski]

Hi Hasna,

Hasna Hocini wrote:
> what I need to know is how I can identify in the SWML file the end of 
> a sentence?

In SWML, punctuation is handled the same as other symbols.  To find the 
end of a sentence, you'd have to find the punctuation that ends a 
sentence.  Most often it would be alone, but I have seen people put 
punctuation with other signs for specific placement.

In the IMWA, the punctuation symbol IDs start with "08-04".  This 
punctuation includes commas, so you'll probably want a specific list of 
BaseSymbols.  Val, can you help us out here?

Regards,
-Steve