Apple’s Spotlight Search gets better at natural language queries in iOS 18

With the Core Spotlight framework, developers can donate content they want to make searchable via Spotlight.
© 2024 TechCrunch. All rights reserved. For personal use only.

Even as Apple embraces AI in iOS 18, it’s working to improve users’ ability to get answers in a more traditional way: through its Spotlight Search feature. Accessed by swiping down on the middle of the Home Screen, Spotlight Search already helps iPhone users discover things on their device and on the web — the latter as an alternative to using Google Search. With iOS 18, Apple is making it easier for people to discover the content inside their apps using natural language — or “semantic” — queries.

Already, Spotlight has been able to search for content in developers’ apps, but it would only return results when search terms matched exactly. With this year’s addition of semantic search capabilities, people will be able to search for content in apps using search terms that are similar in meaning, Apple told developers at WWDC 2024 this week.

With the Core Spotlight framework, developers can donate content they want to make searchable via Spotlight, which is stored in a private index on users’ devices that other apps cannot access. With the new semantic capabilities, Spotlight’s understanding of developers’ content will be much improved as users will be able to search for content inside apps in their own way. These types of search queries leverage machine learning models that are downloaded to the device and run in the app’s process and can be loaded or unloaded at any time, Apple said.

In addition, as Spotlight search is used, developers can help to improve the rankings of their app’s search results over time by sending signals to Spotlight when an iPhone owner engages with one of their results.

For example, if someone were searching for a particular local hiking trail, the semantic search capabilities would be able to find results related to the user’s search query, using entities it understood like the trail’s location or specific keywords, even if they weren’t in the trail’s name and were not exact matches. Plus, the developer can set higher priorities for some of their content, as needed. This allows them to do things like mark a trail that a user had saved as a favorite to rank higher in the search results.

The end result is that the person searches for a given trail using a natural language query in Spotlight, finds the trail they want more easily via these improved results, then can tap on a result to be taken straight to the trail’s page in the developer’s hiking trails app. This update also helps Siri better understand an app’s content, Apple noted.

While not as exciting as an AI-powered Siri or ChatGPT integration, this type of under-the-hood update will make for an improved experience for iPhone users who turn to Spotlight to find information from their apps — or even for those who use Spotlight as an alternative to web search.

 


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *