Google patent suggests ranking content by ‘information gain score’

Image credit: Search Engine Journal

Google may use an “information gain score” to evaluate and rank unique content, according to an analysis of a patent that could influence online publishing and search engine optimization strategies globally.

The patent, titled “Contextual estimation of link information gain,” describes a system for assigning a score to a document based on the amount of new information it provides beyond what a user has already encountered on a specific topic. This mechanism suggests a potential method for Google to identify and reward original content.

First filed years ago, the patent was recently extended in the United States to 2039 and has been cited 24 times, indicating its ongoing relevance to Google’s search algorithms, industry observers said.

Google processes more than 5 trillion searches annually, with approximately 2.92 billion clicks daily directed to the open web, underscoring the vast scale at which content is evaluated.

Experts like Rand, Barry, and Harry Clarkson-Bennett have suggested that Google’s systems likely prioritize originality, potentially through such an information gain score, a Bayesian predictive score, or positive user engagement signals.

The practical application of the patent involves re-ranking or excluding subsequent search results. This process would be based on the novelty and relevance of information provided by a document compared to similar content already presented to the user.

This approach could reduce the visibility of redundant information and elevate content that genuinely adds new insights or perspectives.

The potential implementation of an information gain score means that publishers and content creators might need to focus more intently on producing unique and value-added material to secure higher rankings in Google’s search results.

Source: Search Engine Journal