This is not an easily accomplished feat, as there are a number of moving parts in any web application that make performance monitoring difficult. Oracle database logs do not generally capture performance information, only alerts and errors. The most important question you need to ask is what exactly you are trying to measure: application performance from the user's perspective, performance of the application server (including or excluding database latency), or performance of the database?
Database performance can be affected by a lot of things that have nothing to do with the application, and any given page in the app is likely to be populated by multiple queries, which may be used in turn by multiple forms. If you want to understand each form's affect on the database, I would recommend using the dbms_application_info package to set module and/or action information in the database session from within your application code, and then using a professional performance monitoring tool (not Splunk) to collect and summarize that information so that you can determine each web page's effect on the database and separate out factors not directly related to the application.
If you are talking about monitoring the application tier, then I would re-examine the latency times in your application logs to be absolutely certain they don't include the database latency. Check the Forms documentation or WebLogic documentation to determine what those reported latency times include, and test it as best you can to confirm by comparing browser rendering time with the latency times in the application logs. Splunk is best suited (out of the box) for measuring the application tier's performance.
Client performance can also be affected by a lot of things that don't have anything to do with the application at all, like incidental workstation load, anti-virus scanning, and overall network latency (the load on the network from other users and applications) just for starters. If your end goal is to measure the end-user experience and account for all of those factors then again I would recommend using a tool specifically designed for that purpose and not necessarily trying to customize Splunk for that task.