From: Brandon Amundson ([email protected])
Date: 02/13/02
Hi, (Don't know if this mail will get through to the ml, let's see...). I would like to give a try to the XSB-DAML combination in an attempts to overcome some limitation of our current design for a services platform built around J2EE-SOAP-WSDL and the like. Therefore, I'd need a little help from the more experienced guys in this field, since I didn't find any teaching material on the path to successfull deployment of such a solution. If I'm succeeding, I'd like to share back my results so maybe someone could benefit from it. Today, I've only managed to get XSB and yajxb working, but haven't got too many insights on XSB, or even prolog in general. I've found that Youyong Zou has already written support for DAML in XSB but the missing part is the java program that parses DAML and feeds XSB with the appropriate data. Obviously, I'd love to see the code for such a thing, if it already exists. I'm also wondering if XSB is really considered solid enough to be working in a production environment where it might replace the role of the traditional RDBMS. In particular, how many facts is it able to take into accounts ? And: is the possibility of installing a Oracle or ODBC backend in XSB an option to increase the size of the "facts heap" ? My current idea is the following: Create several DAML ontologies. Have custom GUIs to create the DAML instances. Find a way to automatically generate them from the ontologies. Let users modify the ontologies at will. Setup a DAML crawler to record the location of the DAML instances, as well as to keep track of what instances belong to what ontologies/ontology versions (in case those ontologies evolve). Have the crawler feed an XSB engine with the DAML data. Send queries to the XSB engine. It would be nice to get back the data in DAML for further XSLT processing... Thanks very much in advance, Candide Kemmler --Apple-Mail-3-939161612 Content-Transfer-Encoding: 7bit Content-Type: text/enriched; charset=US-ASCII Hi, (Don't know if this mail will get through to the ml, let's see...). I would like to give a try to the XSB-DAML combination in an attempts to overcome some limitation of our current design for a services platform built around J2EE-SOAP-WSDL and the like. Therefore, I'd need a little help from the more experienced guys in this field, since I didn't find any teaching material on the path to successfull deployment of such a solution. If I'm succeeding, I'd like to share back my results so maybe someone could benefit from it. Today, I've only managed to get XSB and yajxb working, but haven't got too many insights on XSB, or even prolog in general. I've found that <fontfamily><param>Arial</param>Youyong Zou has already written support for DAML in XSB but the missing part is the java program that parses DAML and feeds XSB with the appropriate data. Obviously, I'd love to see the code for such a thing, if it already exists. I'm also wondering if XSB is really considered solid enough to be working in a production environment where it might replace the role of the traditional RDBMS. In particular, how many facts is it able to take into accounts ? And: is the possibility of installing a Oracle or ODBC backend in XSB an option to increase the size of the "facts heap" ? My current idea is the following: Create several DAML ontologies. Have custom GUIs to create the DAML instances. Find a way to automatically generate them from the ontologies. Let users modify the ontologies at will. Setup a DAML crawler to record the location of the DAML instances, as well as to keep track of what instances belong to what ontologies/ontology versions (in case those ontologies evolve). Have the crawler feed an XSB engine with the DAML data. Send queries to the XSB engine. It would be nice to get back the data in DAML for further XSLT processing... Thanks very much in advance, Candide Kemmler</fontfamily> --Apple-Mail-3-939161612--
This archive was generated by hypermail 2.1.4 : 03/26/02 EST