Build your First Semantic AI Search Application with the PoolParty Application Development Framework

The PoolParty Application Development Framework (ADF) is a low-code/no-code platform that slashes development time for Semantic AI applications from months to hours. It provides a comprehensive toolkit, including workbenches, pre-built frontend applications and APIs, to support you at all stages of your journey – from initial POC to final integration.

So, let’s get started! In this tutorial, you’ll learn how to leverage the PoolParty Application Development Framework to build, test and integrate Semantic AI search.

Prerequisite

Before you start, make sure that you have the following items ready:

• PoolParty taxonomy
• Suitably configured GraphSearch search space
• PoolParty login credentials permitting UI and API access to ADF

Step 1: Build ⚒️

Log in to the Administration backend of the Search Application and create a new configuration. The configuration will hold your custom settings. Type in the name of your configuration and select a GraphSearch search space.

If you want to build a search application with facets, it is very important that the search space you select has a pre-configured facet model. To verify that everything works as expected, navigate to the Filter settings tab. You will see that the fields are populated with the concept schemes from the connected taxonomy.

For the sake of this tutorial, it’s fine to leave all other settings at their defaults. If you are curious about the meaning of the individual fields, just hold your pointer over the Help icon and a tooltip pops up.

Once done, confirm with Save. Your settings have been saved. Click the PoolParty logo in the top left corner to navigate to the main user interface.

Step 2: Test the Application 🔬

Depending on your organization’s needs, the main user interface of the Search Application can serve as a ready-to-go search application, as a demo interface to explore the capabilities of semantic AI or (as in our case) as a test environment where we can verify that everything works as expected.

First select the configuration you created in the previous step. 

Now go ahead and start typing your search query into the search box. ADF supports both concept search and full text search. If your search query matches the preferred label of one of the concepts from the connected taxonomy, you will be able to select that concept from among provided suggestions.

You can also try to narrow down your search results with facets available on the left hand side or concepts associated with the returned documents.

Step 3: Integrate 🧩

To interact with the Search Application programmatically, you’ll need to authenticate your requests with an OAuth 2.0 access token. To get the token, send a request with your username and password to Keycloak, a third-party component PoolParty (and by extension also ADF) relies on for user management.

In ADF APIs, configurations are not identified by their name but by their ID. To find out the ID of your configuration from the previous step, send a GET request to /api/configurations. A successful call will return settings and metadata of all ADF configurations together with their IDs.

The configuration settings also come handy if you want to give your users an option to narrow down the search results with search filters – the facetsToShow property contains all facets available for a given configuration.

If you want to help your users to complete their queries faster, you might want to make use of the /api/configurations/{configId}/suggestConcepts endpoint. This endpoint will return concepts matching the user’s input and thus add an autocomplete functionality to your semantic AI search application.

Finally, to get search results, send a POST request with the search query to /api/configurations/{configId}/search.

Excellent work! In this tutorial, you built your first Semantic AI search application using the PoolParty Application Development Framework (ADF). You learned how to create and test a new configuration using the Search Application UI as well as how to integrate the ADF functionalities into your custom application using the provided APIs.