Hi Enzo,
Your english is perfect!
First of all, there is a big chance, that the name in the text is
recognized as a Person. Several different approaches are used for this
purpose. First, there may be an exact match of course with a known
person from the KB (Knowledge Base). Several other mechanisms are
triggered too , but the exact match from Person in the KB has the
biggest weight. KIM also contains a list of well known name parts like -
first name, surname etc. which it uses in jape rules to recognize
Persons. The list is contained in the PersonLexica part of PROTON and
the KB. So if a part of the name is matched as an instance from the
PersonLexica, it's almost certain that a Person annotation will be
created. KIM also uses job titles, public titles, name prefixes (al-),
initials etc. for the recognition. You can look at some of the japes
like firstname.jape, person_name.jape etc. and it will become more clear.
These rules will also help you construct your custom rules to recognize
SELLERs.
all the best,
Philip
ipinkfl...@inwind.it wrote:
Hi,
my name is Enzo Buono, and I'm an Italian student in Computer Science. Sorry
for my English.
I have a question about KIM. I need to use information extraction. I have a model of legal
ontology. In this ontology there is the concept of SELLER (There are other classes, I use this for
my example). I would like to create a system which would be able to recognize the concept of
SELLER, keeping in mind that KB doesn't contain the corespondent instance. Now i'll give you an
example to explain what i want to say. I suppose that is possible to recognize the entity of SELLER
by the use of a rule like"SELLER IS A PERSON IF IN THE SAME PHRASE THERE IS THE VERB
"SELL" ".
So i want to know if it is possible to recognize an instance without the consultation of KB prepopolated ,but by the use of a rule...
Thanks and regards,
Enzo
_______________________________________________
Kim-discussion mailing list
Kim-discussion@ontotext.com
http://ontotext.com/mailman/listinfo/kim-discussion
_______________________________________________
Kim-discussion mailing list
Kim-discussion@ontotext.com
http://ontotext.com/mailman/listinfo/kim-discussion
