From aaacf84cee6976fadb5deb9cb8705914d705e09f Mon Sep 17 00:00:00 2001 From: Michael Kleber Date: Fri, 2 Apr 2021 12:42:46 -0400 Subject: [PATCH] Mention cohort stability in FLoC README As @AramZS points out, #58 is a recurring question that deserves some mention here. --- README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/README.md b/README.md index f967940..a6e48c7 100644 --- a/README.md +++ b/README.md @@ -30,6 +30,8 @@ creative = await fetch(url); The browser uses machine learning algorithms to develop a cohort based on the sites that an individual visits. The algorithms might be based on the URLs of the visited sites, on the content of those pages, or other factors. The central idea is that these input features to the algorithm, including the web history, are kept local on the browser and are not uploaded elsewhere — the browser only exposes the generated cohort. The browser ensures that cohorts are well distributed, so that each represents thousands of people. The browser may further leverage other anonymization methods, such as differential privacy. The number of cohorts should be small, to reinforce that they cannot carry detailed information — short cohort names ("43A7") can help make that clear. +The meaning of a particular cohort should stay roughly consistent over time. As individual people's browsing behavior changes, their cohort will change too, but the algorithm that turns input features into cohort assignments should remain stable. If that cohort assignment algorithm does eventually need to change, then the migration to a new assignment algorithm will need to be clearly communicated by the API, so that consumers of the cohort signal are well informed of the need to update their usage. (See [Issue #58](https://github.com/WICG/floc/issues/58) for more on this topic.) + ## Privacy and Security Considerations There are several abuse scenarios this proposal must consider.