Seevl, the music discovery service built on Semantic Technology that I wrote about a few months ago, has released a significant update to their plugin for YouTube. The plugin is still only available for the Google Chrome browser, but other browser plugins are in the works. You can grab the Chrome plugin here.
Once the plugin is installed, the user has new options available when visiting YouTube. First, there's a new search option next to the standard YouTube search bar.
That link adds semantic search functionality to YouTube. For example, a user could select 'Origin = Los Angeles' and 'Genre = Cuban Jazz' and discover Luis Bonilla. In fact, Seevl allows for up to 10 criteria to identify bands /
As Seevl CEO Alexandre Passant says, "users can find music such as "Alternative rock bands from California" or "Anyone started to play electro in 1984", going further than plain-text search capabilities of YouTube music
While the user is directly performing the query, and the results are shown in the YouTube interface, the back-end calculation is being performed using seevl search capabilities (which means, SPARQL queries on top of the seevl
knowledge base, rendered as JSON-LD).
But that is just where the fun starts. On the individual page for a video from the L.A. band, Ozomatli, the Seevl plugin displays other faceted views of information related to the artist: Biography, Fact-Sheet, and Suggestions. These are pretty self-explanatory, and I covered them briefly in my earlier post, but it's worth calling out a bit of the functionality.
In the Fact-sheet, each data point provides a way to drill down into the data collected by Seevl. Want to know about individual band members? What about particular genres the music's been categorized into and other artists who play that style? It's all there.
But if you'd like to have Seevl take the extra step of serving as a recommendation engine, "Suggestions" will do this.
Suggestions was available when the plugin originally launched, but what's new here is the addition of a layer of explanation. Why did Seevl suggest a particular artist? Clicking on the puzzle piece reveals the various connections between the artists.
There is some room for improvement in the plugin, and the algorithm seems a bit buggy at times, but the pieces are in place for a very powerful music discovery service for YouTube and other existing platforms. In fact, that model is exactly what Passant has in mind: "Based on feedback so far, we decided to concentrate our efforts into integrating our feature into existing apps, so that people can discover music directly into their favorite services. So adding these new features to the YouTube plug-in made sense, and we decided to launch the new website at the same time to make the message clearer."