Topics EE - GoogleChromeLabs/ps-analysis-tool GitHub Wiki

Explorable Explanation: Topics Timeline

The Topics API is a privacy-preserving technology developed by Google as part of its Privacy Sandbox initiative. It aims to replace traditional cookie-based tracking methods, which have raised significant privacy concerns.

The Topics API works by analyzing a user's browsing history and identifying general topics of interest. It then assigns a few of these topics to the user for a week. When the user visits a website, the browser shares one of these assigned topics with the website. This allows the website to show relevant ads without knowing the user's specific browsing history.

Taxonomy in Topics API

The Topics API uses taxonomy, which is essentially a hierarchical classification system, to categorize user interests. This taxonomy consists of a set of predefined topics, such as "Technology", "Travel", or "Fitness."

Understanding Topic Assignment in the Explorable Explanation

The Explorable Explanations provide a visual representation of how the Topics API assigns topics to users based on their browsing behavior.

Key Concepts:
  • Epoch: An epoch represents a specific time period, typically a week. During each epoch, a user is assigned a limited number of topics.
  • Node: Each node in the visualization represents a website.
  • Colored Circles: The colored circles attached to a node indicate the topics that are assigned to the user when they visit that particular website during a specific epoch.

How Topic Assignment Works:

  • Topic Generation:

    • The browser periodically analyses the user's recent browsing history.

    • It identifies general topics or interests based on the visited websites

    • These topics are categorized into predefined categories like "Fitness," "Technology," or "Travel."

  • Topic Assignment:

    • A limited number of topics are assigned to the user for a specific epoch.

    • This information is stored locally on the user's device.

  • Topic Sharing with Websites:

    • When the user visits a website, the browser randomly selects one of the assigned topics for that epoch.

    • The selected topic is shared with the website.

Visualizing the Process:

The visualization in the Explorable Explanations helps to understand this process:

  • Node Colors: The colour of a node indicates the topic assigned to the user during a specific epoch.

  • Multiple Colors: If a node has multiple colors, it means that the user has been assigned to different topics during different epochs, depending on their browsing behavior.

  • Topic Selection: When the user visits a website, the browser randomly selects one of the colored circles attached to the node. The colour of the selected circle represents the topic that is shared with the website.