Microsoft has launched a patent named Net Content material Reliability Classification that talks about tips on how to develop a reliability rating for an internet site or content material on an internet site. The patent looks like it may be utilized by the Bing Search group for higher rating of internet sites and net content material, however that doesn’t imply it’s at present getting used within the Bing Search outcomes at present.
The patent was revealed on November 2, 2023 after being filed on July 5, 2023 – you may read it over here.
Highlights. Listed below are some attention-grabbing highlights from this patent utility.
- The reliability rating can be utilized to dam content material, rank content material, present a content material warning, and choose a supply to reply a query, together with different makes use of.
- Visitors information can point out whether or not a supply is in style, however in style just isn’t the identical factor as dependable.
- Pure language processing can be utilized to find out whether or not on-line content material is grammatical, however grammatical can also be not the identical factor as dependable.
- The current expertise identifies dependable content material by leveraging professional scoring for a small quantity of net content material by iteratively extending these scores to different content material primarily based on how net content material is linked.
- Consumer interactions can also be leveraged for figuring out a reliability rating as effectively
- The excessive reliability rating is generated by first figuring out excessive reliability on-line content material inside an internet graph.
- These initially scored websites could also be described as seed websites.
- Scores for the seed websites could also be taken from authoritative lists of identified dependable content material suppliers
- An output of the expertise is a excessive reliability rating and a low reliability rating for an internet content material.
- Totally different purposes can eat this rating to carry out or information completely different capabilities, together with search, filtering, content material warning technology, and the like.
The summary. Right here is the summary of the patent:
Know-how described herein assigns a reliability rating to net content material, reminiscent of a website online or portion of an internet site. In a single facet, an output of the expertise is a excessive reliability rating and a low reliability rating for an internet content material. The excessive reliability rating represents conformance to excessive reliability websites, whereas the low reliability rating represents conformance to low reliability websites. The excessive reliability rating could also be generated by first figuring out excessive reliability on-line content material inside a compressed net graph. In a primary iteration, the excessive reliability rating of the seeds is used to attain on-line content material that’s linked to the seed websites. At a excessive stage, the extra hyperlinks that originate from excessive reliability sources, the upper the reliability rating for the linked content material. The low reliability rating is analogous, however makes use of outgoing hyperlinks to low reliability websites as an alternative of incoming hyperlinks from excessive reliability websites.
Why we care. Many SEOs take pleasure in studying patent paperwork from the Google and Bing Search groups. Whereas we all know that simply because a patent has been filed, it doesn’t imply a search engine is utilizing the expertise as described within the patent within the dwell search outcomes. Both approach, it may be instructional and helpful to grasp how these search scientists who work at Google and Bing take into consideration these rating and scoring challenges.
Hat tip to Glenn Gabe for recognizing this patent.
New on Search Engine Land