Have you taken a look at Apache Lucene?
On 10/20/2010 03:53 PM, Pedro wrote:
It helps a little bit, thank you for your prompt response.
I am trying to rank the web services according to a query, like in
information retrieval. However, I guess that the queries are based on
key-words making a matching with the information provided in natural
language for each web service, or based on key IDs, getting directly
the structures which have that ID.
My real doubt is: suppose we have a weather web service, that has one
operation, and this operation receives as input a ZipCode, and as
output the Temperature from that place. I'd like to know if there is
any way of using a query to specify some sentence like that in UDDI
(show me all the services that have ZipCode as input, and Temperature
as output) and then rank the services according to the similarity to
the specified query.
I hope my doubt was understandable.
Best Regards,
Pedro
Em 20/10/2010 17:20, Kurt T Stam escreveu:
Hi Pedro,
1. What is it you are trying to rank?
2. UDDI is like a yellow pages, meant for looking up WebServices and
for storing references to additional information
about them. Additional Info could be anything, for example WSDL,
Input/Output Documentation. But the key word in this sentence is
*reference*. A repository would store the Docs, a (UDDI) registry
points to it.
Does this help?
--Kurt
On 10/20/10 2:59 PM, Pedro wrote:
Hi jUDDI Folks,
I am working on Semantic Web Services, and I need a baseline to
compare my experiments. I've installed jUDDI, published some
services on it, read some stuff about UDDI, and I guess UDDI is not
capable of ranking web services from a query, and a query in UDDI is
not capable to specify details as web services inputs and outputs.
I'd like to know if it is really what happens or I am just saying
nonsense.
Best Regards,
Pedro
