BY THE YOUTUBE TEAM
Starting today, we are testing an experimental feature to allow people to add notes to provide relevant, timely, and easy-to-understand context on videos. For example, this could include notes that clarify when a song is meant to be a parody, point out when a new version of a product being reviewed is available, or let viewers know when older footage is mistakenly portrayed as a current event.
This work builds on a number of products we’ve launched to display helpful context alongside videos, such as information panels, or our recent disclosure requirement when content is altered or synthetic.
The pilot will be available on mobile in the U.S. and in English to start. During this test phase, we anticipate that there will be mistakes – notes that aren’t a great match for the video, or potentially incorrect information – and that’s part of how we’ll learn from the experiment. Viewers, participants, and creators are invited to give us feedback on the quality of notes, and we’ll look at ways to improve over time, including whether it makes sense to expand the feature.
Who is writing and rating notes?
To start, a limited number of eligible contributors will be invited via email or Creator Studio notification to write notes so that we can test the feature and improve our systems before we consider expanding. Eligibility criteria include having an active YouTube channel in good standing with our Community Guidelines.
Viewers in the US will start to see notes on videos in the coming weeks and months. In this initial pilot, third-party evaluators will rate the helpfulness of notes, which will help train our systems. These third-party evaluators are the same people who provide feedback on YouTube’s search results and recommendations. As the pilot moves forward, we’ll look to have contributors themselves rate notes as well.
How does note rating work?
Notes will appear publicly under a video if they’re found to be broadly helpful. People will be asked whether they think a note is “helpful,” “somewhat helpful,” or “unhelpful” and why – for example, whether it cites high-quality sources or is written clearly and neutrally.
From there, we’ll use a bridging-based algorithm to consider these ratings and determine what notes are published.
A bridging-based algorithm helps identify notes that are helpful to a broad audience across perspectives. If many people who have rated notes differently in the past now rate the same note as helpful, then our system is more likely to show that note under a video. These systems will continuously improve as more notes are written and rated across a broad range of topics.
People come to YouTube to share and learn about the world around them. With this feature, we’re excited to explore new ways to make context even more relevant, dynamic, and unique to the videos you’re watching, at scale across the huge variety of content on YouTube. We look forward to getting feedback from viewers, contributors, and creators in the weeks and months ahead as we work to support information quality overall on the platform.
(Source: YouTube)
