/Patent Tracker
MicrosoftPatent applicationAI & softwareInterest 6/10

Microsoft patents method for feeding web page structure to language models

Patent drawing
Patent drawing from the official filing.

Microsoft filed a patent application describing a method to extract structural and content data from a web page's accessibility information, filtering out visual display details like font styles. This data is converted into markdown format and selected based on relevance to provide context to a language model, helping it generate more accurate responses about the page's content.

Why it mattersThis approach could make AI assistants better at understanding and summarizing web pages by using the underlying structure of the content rather than just the visible text.

In the filing’s own words

Some embodiments get accessibility information of a web page, and then filter it to include specified context data and to exclude specified non-context data. A version of the context data is provided to a language model in a prompt, or along with a prompt, to influence computation of a response of the language model. In some cases, the provided version includes a markdown language translation of accessibility data extracted from the web page. Some embodiments run a browser script before extracting the accessibility data. In some, a portion of the markdown is chosen, with the choice being based on vector similarity or on association with a link. In some embodiments, links are consolidated.

What exactly they claim (claim 1)

1. A method performed by a computing system, the method comprising: extracting accessibility data by extracting structure data or content data, or both, from accessibility information of a web page, the extracting comprising filtering such that the extracted accessibility data is free of text display data which specifies a visual display characteristic of a text string; translating at least a part of the extracted accessibility data into a markdown language text; and supplying at least a portion of the markdown language text to a language model as a prompt context.

Facts

Company
Microsoft (Microsoft Technology Licensing, LLC)
Official title
ACCESSIBILITY DATA SELECTION WITH PROMPT CONSTRUCTION
Publication number
US20260310743A1
Application number
19169051
Filed
Apr 3, 2025
Published
Oct 8, 2026
Inventors
4
Claims
20

Classified under

G06F 16/986 Physics & computingG06F 16/951 Physics & computingG06F 40/58 Physics & computing

More from Microsoft in AI & software

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.