Hi.
While creating a Corpus() object whith "Corpus c = new
Corpus("myCorpus.xml")", I allways get a
java.lang.OutOfMemoryError.
I use the original Tiger Corpus from the homepage not
just the sample. (No Error occured while loading the
sample file.)
What can I do to avoid this error? I hope to receive
some hints.
ps: I've allready tried to avoid the problem by increasing
the stacksize.
Logged In: YES
user_id=947333
You have to split the corpus into smaller units. Loading the
whole TIGER corpus (~40000 sentences) will require about
1.4 GB of memory. So, on a machine with 512 MB memory
available it should be sufficient to split the corpus into 4
parts - you can use a text editor to do this; make sure that
all parts are well-formed XML. Then start your application
using the -Xmx500M switch.
Example: java -Xmx500M -cp tigerAPI.jar sampleApp
Please let me know whether it worked.
oeze