Google has recently been granted a patent for ranking web pages, shedding light on how AI Overviews ranks content. The patent, titled “Contextual Estimation Of Link Information Gain,” was filed in 2018 and awarded in June 2024. It outlines a method for ranking pages based on what a user may be interested in next.
The patent describes a scoring system called Information Gain, which determines the likelihood that a user will find a second set of web pages relevant as a slightly different follow-up topic to a previous query. Think of it like building a pizza – starting with a base, adding toppings, and creating layers of specifics to form the complete context.
Contrary to common misconceptions, Information Gain is not about boosting the ranking of web pages with more information. Rather, it focuses on understanding how AI Overviews ranks web pages that a user may be interested in next. This patent introduces a new dimension in how web pages are ranked, particularly within the context of automated assistants or chatbots.
The emphasis of the patent lies on automated assistants, with references to search engines playing a supporting role. The goal is to anticipate what a user may be interested in next and provide relevant information accordingly. The patent emphasizes the importance of presenting information in a concise, accurate, and relevant manner to enhance the user experience.
One key aspect of the patent is the concept of Dialog Turns, referring to the back-and-forth interaction between a user and an automated assistant in a conversational manner. By understanding the additional value that a new document provides compared to what the user has already seen, the patent aims to improve efficiency and relevance in delivering information.
The patent also addresses the issue of information density, particularly in audio responses, where long-winded answers may not be ideal. By focusing on providing relevant and concise information, the patent aims to streamline the user experience and avoid redundancy in responses.
While the patent does mention a search results interface for presenting the second set of ranked web pages, it is important to note that the focus is not on organic search results. Instead, the patent is centered around presenting documents within the context of an automated assistant or chatbot.
In conclusion, the Information Gain patent offers valuable insights into how web pages are ranked based on user interests and context. By leveraging this scoring system, automated assistants can anticipate what users may want to see next and provide relevant information accordingly. This patent represents a significant development in the realm of AI Overviews and automated assistants, enhancing the user experience and efficiency in accessing information.
For further details on the Information Gain patent and its implications for web page ranking, refer to the latest version of the patent: “Contextual Estimation of Link Information Gain.”