G-NRIS stands for Graph Neural Reader Intelligence System. It is the intelligence system behind many of the ways bks.social understands books, readers, conversations, and discovery.
Traditional systems are linear and put readers into buckets: You liked Book A, so you MUST like other books like Book A. Eventually, the entire social feed reinforces that assumption, becoming less diverse and pigeonholing both readers and content creators.
G-NRIS doesn't assume reading taste works that way. The system understands that readers can love different books for different reasons and that's why G-NRIS was designed to map why readers connect with books by looking at the relationships between books → readers → authors → genres → themes → tropes → reading preferences → and conversations.
The G-NRIS system analyzes/collects information about books via signals and affinities. Over time, those connections help bks.social build a better picture of both the books in the community and the readers who enjoy them.
G-NRIS considers literary characteristics such as genres, themes, tropes, moods, and other story elements rather than treating a book as only a title in a category.
Reading activity and conversations can reveal patterns in what a reader responds to, helping build an evolving picture of their reading interests.
Those connections help bks.social surface books that make sense for a particular reader instead of relying only on popularity or broad genre matches.
G-NRIS connects books with the discussions, reviews, creators, readers, and discoveries around them so useful literary context does not have to disappear with the feed.
G-NRIS does not treat its understanding of a reader or a book as permanent. As more reading activity and conversation happen on bks.social, the system can refine the connections it has already made and recognize new ones.