Discovery Explainer Now Available to All Users
Leveraging Large Language Models (LLMs) to instantly understand insights
After launching in Early Access this summer, we are now officially rolling out Discovery Explainer to all Radarly users!
What is it?
Discovery Explainer is a new feature in Discovery that uses Large Language Models (LLMs) to contextualize and bring to life emerging terms found by Discovery. It scrolls, analyzes, and understands the posts behind emerging terms in order to generate a human-like text explanation that allows you to get context about about the emerging terms it has identified inside social data.
This feature will be revolutionary to instantly understand insights, and to further democratize the data within your organizations.
What’s the value?
- Discovery Explainer shortens the time to insight by not just detecting, but also processing and explaining the the social media learnings it identifies in users’ language.
- By harnessing the power of LLMs, this feature unlocks a deeper understanding of consumer intelligence data, enabling businesses to stay ahead in a competitive market.
- No need to know the queries of a Workspace inside and out - by providing the context around an emerging term, this feature empowers non-analysts and other occasional users to use and understand Discovery insights.
- Discovery Explainer leverages the power of Generative AI to provide context and understand how the software identifies and displays learnings.
How does it work?
Once requested by the user, Discovery uses a Large Language Model (LLM) to scroll, analyze and understand the posts behind the detected peak, and generates a human-like text that allows you to better understand what’s happening behind the data.
This feature allows users to understand the context of social media trending terms, like:
- Key phrases
- Hashtags
- Affects
- Emotions
- Sentiment
- Content Classification (Last 7 days only).
Explanations will be delivered in the language you have chosen for the Radarly UI (English, Chinese, French, Japanese, or German).
Learn more in the Help Center:
Post Geolocation Detection Based on Images
What is it?
New AI capability that allows automated detection of famous places, cities and countries in images contained in posts.
When a country, a city, or a place is detected in an image, this geolocation is automatically attributed to the related post (if the post has not already been geotagged by the author). This capability complements the other methods of post geolocation.
What’s the value?
This new image analysis capability allows you to:
- Benefit from a better precision on countries and cities detection to produce analysis and statistics
- Filter posts related to an emblematic place for qualitative analysis