From technology down to Python, Anime & manga down to Pokemon, social media giant Twitter offers direct access and interaction with a diverse variety of topics and topic expertise authorities.
The site’s help center states, “we want Topics on Twitter to reflect the broader, lasting conversations people have about the events, people, and things they discuss. So we use machine learning to find related Tweets from these conversations. This could mean they Tweet a lot about the Topic or interact a lot with Tweets about the Topic. From there, we find the Tweets that are most interesting to those people, using algorithms, keywords, and additional signals.”
Building upon their current topic processes, Twitter has been issued a patent for a process that will help the platform identify and rank authorities for an expertise topic.
About the Patent Application
Twitter’s patent describes an iterative process that can be used to determine the subject matter expertise of users on a social network. User accounts are analyzed by how frequently they interact with a plurality of topic groups (TGs). As users are identified to frequently interact with multiple topic groups they are continually filtered to more specific TGs. After a certain amount of iterations, a ranking can be obtained of accounts that are an authority on the specific expertise topic.
Twitter’s invention has the capability to organically determine subject matter experts of a particular topic based on the way they interact with a social network. Through the use of an iterative grouping and analysis process, large social networks can determine the users most qualified on specific topics.
In one account of the detailed description, the patent describes:
“…Accounts which are consistently found to be members of the same topic groups in which the top accounts are found are identified. For example, in the technology expertise topic, there may one or more accounts which appear as members of the topic group in which all three accounts named Bill Gates, Tim Cook, and Elon Musk appear, but who are not yet identified as authorities on the expertise topic. This account is then added to the authorities ranking being compiled… and this identified account may be ranked as the next prominent account for the expertise topic group after Bill Gates, Tim Cook, and Elon Musk, as an authority on the expertise topic.”
“Alternatively, any random topic group in which at least three of the top ranked accounts appear as members may be processed to determine the next frequent account that coexists with those of Bill Gates, Tim Cook, and Elon Musk. For example, the next account that coexists with Bill Gates, Tim Cook, and Elon Musk in membership of the topic group with an expertise topic ‘computer technology’ which may be a child node of the expertise topic ‘technology’ may be Mark Zuckerberg. In this case, Mark Zuckerberg is identified as appearing as a member of the technology topic group… and this account is added to the top authorities on the expertise topic ‘technology.'”
Additionally, when it boils down to how many topics a user can be an expert on, the patent hints that “each account may only have up to a maximum number of associated expertise topics, e.g., two per account. “
why This Patent Application is Interesting
The method disclosed in the patent is very similar to traditional digital marketing techniques in the way that the user analysis and grouping are performed. However, what makes this patent unique is how the traditional user profile analysis focuses on the subject matter a user interacts with to determine their expertise on the subject, rather than their marketability to an item. The iterative approach enables this to be applied to all fields, as the system will be able to internally refine the TGs that it processes.
Written by John DeStefano, Technical Advisor
and Lauren Hawksworth, Marketing Administrator
March 30, 2021