Algorithm: There are a lot of books in the library that I have not read. I want to use some method to calculate the hypothetical sorting result-suppose I have read all the books.

Algorithm: There are many books in the library that I have not read, I want to use some method to calculate the hypothetical sorting result. suppose I have read many books in all the bookstores that I have not read, I want to use some method to calculate the hypothetical sorting result-assuming that I have read all the books (of course I have scored points on these books ). In this way, I can select a better algorithm for searching books based on the sorting results: there are many books in the library that I have not read, I want to use some method to calculate the hypothetical sorting result-assuming I have read all the books
There are a lot of books in the library that I have not read, and I want to use some method to calculate the hypothetical sorting result-assuming that I have read all the books (of course I have scored points for these books ). In this way, I can select a book that may be better for me based on the sorting results.

I have an idea, but I don't know why or how to implement it. If we can find someone who is "different" (hypothetical person with the same hobby, taste, level, etc., in short, we both score the same for any book ), enough "different" books I have read
It should cover all books. That is to say, if I have read all the books and have scored them, I can sort the scores to get the desired results.
Of course, this is a bit idealistic for situations that do not limit the scope of the book, because it may not be able to find enough "different" me. For example, I want to find my team in basketball, football, rugby .... people with the same preferences (those who have the same preferences for these teams) may be difficult to find people who have the same preferences for basketball. Therefore, we must limit the scope of books before calculating the hypothetical sorting.

But how can this idea be implemented? Or who can provide other ideas and implementations that meet the requirements? Algorithm ???? Sort and share:
------ Solution --------------------
Simply put, your idea is to partially deduce the entire set, which is both reliable and unreliable.

Reliable means that many people are doing research in this area, which shows that your ideas are correct.
Unreliable analysis refers to the absence of authoritative papers (which is not included in the single-factor analysis)

There are quite a few mathematical models, such as neural systems, Logistic regression, and progressive chromatography ...... And so on.

My first job was to do something similar-Market Research
Determine the development direction (product attributes) of new products (that is, unknown products) based on the feedback from the sample population)

If it's just a single-factor linear infinite value, you can usually find development goals in its direction. for example, if you are talking about a certain sport, you only prefer it and do not like it most, find people of the same kind in the direction you like or hate

The above is only a perceptual discussion. I am afraid I have to search for documents and create a mathematical model by myself.