Search This Blog

Showing posts with label PLSQL. Show all posts
Showing posts with label PLSQL. Show all posts

Thursday, March 13, 2025

Leveraging the Oracle Autonomous Database and AI for Effective Content Moderation in Online Communities

Today, managing online communities and forums can be a daunting task. From duplicate questions to unaccepted answers, and from unchanged idea statuses to duplicate ideas, the challenges are numerous. Even with active management, the manual effort required by product teams can be overwhelming. However, AI-driven content moderation offers a promising solution to these pain points. In this blog post, we'll explore how AI can be used to streamline content moderation and enhance the user experience in any online community.

Addressing Duplicate Content

One of the most common issues in online communities is the presence of duplicate ideas or questions. Here's a methodology to tackle this problem using the Oracle ADB and Gen AI Service in OCI:

  1. Database Integration: Connect your community's database to an AI service. For instance, if your community is hosted on an autonomous database (ADB), you can set up a trust between the ADB and an AI service (like the OCI Gen AI Service) natively.
  2. Creating Views: Develop views that represent the community categories, containing idea numbers, names, and descriptions, for example, as metadata for the AI Service to utilize.
  3. AI-Powered Duplicate Detection: Create a stored procedure (API) that calls the AI service to identify similar ideas/questions already created. Based on a predefined accuracy/confidence level, the AI can flag potential duplicates.
  4. Automated Comments: Develop an API to automatically comment on flagged ideas, informing users about the duplication and providing links to similar ideas or threads.
  5. User Feedback and Model Training: Allow users to challenge the duplicate flag. These challenges can be used to train a custom model over time (you can start with Oracle's AutoML feature with very low effort), improving accuracy with minimal effort. This would be done at the client end.
  6. Nightly Jobs: Schedule nightly jobs to run within the database to perform this true-up, using AI services and APIs to perform regular content moderation.

Enhancing Search Validation

To further strengthen content moderation, consider the following steps:

  1. REST Enablement: REST enable the stored procedure API using a service like Oracle Rest Data Services (ORDS), which is native within the ADB.
  2. Improved Search Validation: Integrate this API as a validation step when users create new ideas or questions in your website or app. This step can enhance the existing search validation by considering different languages, descriptions, and metadata, ensuring a more accurate detection of duplicates.

Flagging Existing Capabilities

Another opportunity for AI-driven content moderation is to flag ideas/questions/enhancements that suggest features already available in your product:

  1. Repository Integration: Expose a repository of current capabilities to the AI service.
  2. Automated Comments: Develop an API to automatically comment on flagged ideas, providing links to documentation and informing users about existing features.

Conclusion

AI-driven content moderation offers a powerful solution to the challenges faced by online communities, including for product teams that crowdsource content to improve their software solutions. By leveraging AI, communities can reduce manual effort, enhance user experience, and ensure more accurate and efficient content management. Whether it's detecting duplicate content, improving search validation, or flagging existing capabilities, the ADB and OCI Gen AI Service can transform the way online communities are managed, driving growth and success.

Sunday, May 24, 2020

EBS Integration Repository and eTRM


As a lot of companies and users shift their focus towards the Cloud Products, sometimes it's easy to forget that an enormous number of customer still utilize EBS and will continue to do so for years to come. As such, those customers will continue to integrate their EBS environments with other systems, like early payment engines, OCR engines, tax engines, third party SCM systems, other ERP's and more. The following two resources are extremely powerful when planning integration work for EBS, and I have leveraged them countless times.

First, I wanted to remind everyone that the Integration Repository responsibility in EBS is a great resource to see whether there is already an Oracle API to accomplish a specific task. Please make use of it when building solutions to address customer needs or if you have doubts in regards to whether the vendor you are working with for your project, or your own development team, is building a custom function for an existing standard function or procedure, as you always want to leverage seeded functionality that will be supported by Oracle and doesn't break when upgrades happen.

For those that are not familiar with the Integration Repository, you simply need to add it as a responsibility under your user and browse the different modules to drill down to the specific service interface you want to look at. Once you drill into a specific API, you will see a list of the details including function names, parameters, rules, and much more.

Browsing the repository:



  • I am showing Projects related API's in the screenshot.

Additionally, remember the Oracle eBusiness Suite Electronic Technical Reference Manual (eTRM), is also available to us with our Oracle Support accounts.

This one is especially useful to see EBS database design and dependency information, you can also browse the FND model and the Data Dictionary with it! This is an extremely powerful tool to see what happens to those database objects that you rely on when upgrades occur, as those changes will likely impact your custom concurrent programs, reports, alerts, and other extensions.