PROBLEM with Non-Fiction Book/Author Rating/Stats/Name Variation & Potential Solution

[mikhail_fil]

I use Calibre to organize a lot of science books and journal papers, as I am sure so many other people do.

A lot of these books and papers, tend to have multiple authors that either collaborated on the same book or paper, or have worked on different chapters of the book, thus a book/paper may have been written by 20-50 contributors. If I assign 5-star rating to this book, does the rating apply to all authors? If yes, it fills the library with misinformation, if there is some type of algorithm, well, same problem.

Star Rating is good for books with 1-2 writer max who worked together. Otherwise, its almost useless, unless the library contains a lot of curated single author entries.

It makes me more difficult to make good "connections" between the authors working together, because its hard to quickly identify "rising stars" out of the endless pools of 1-hit students on teams of 50.

Phenomenal Solution:

Query Google Scholar Stats on Author and/or paper. These stats are all public, and design of google scholar is unlikely to change much. Once the stats are downloaded for an author, Calibre could simply confirm the author with User for any new entries and only download citations for the new paper/book.

Having this information in Calibre database would actually create tremendous value by helping to identify papers by hard working researchers that may not have an instant name recognition, and would really help to "follow the thread". It would really be a big game changer in managing nonfiction work.

my library has over 20 thousand authors, most of the names are meaningless to me, its nearly impossible to manually search and update this type of information. Don't even get me started on similar/same names which mess up the count of books/papers per author.

By doing a single search on google scholar with Name and Paper/Work to establish a connection, we can link each author to his/her Google Scholar profile. Pull the H-Index, total citation count per author, and citation counts for individual publications.

(an example of Karl Frison's Google Scholar Page - https://scholar.google.co.uk/citatio...0aoAAAAJ&hl=en)

This can and should lead to further possibilty to setup author publication alerts on Calibre, as well as direct downloads from Open Access sources. I know that Calibre is free, but I would eagerly pay a monthly fee to avoid the manual addition and metadata updates that waste so much time. Or instantly pull up the profile of an author by clicking the author's name. Or seeing which papers I have yet to download for the author. Then assign my own rating. There are endless possibilities, and they don't seem too complex nor do they seem to rely on paywalled data much.

I would love to hear thoughts and ideas. (no, zotero and mendley are not good for this, mendeley is only good for citations and abstract downloads, and you are forced to keep your books separately in calibre, creating a search and stats nightmare.

[kovidgoyal]

authors dont have a rating. when you rate a book that rating applies to the book only. The rating that is used to color author icons in the tag browser is the average rating of the books by that author.

[mikhail_fil]

Originally Posted by kovidgoyal

authors dont have a rating. when you rate a book that rating applies to the book only. The rating that is used to color author icons in the tag browser is the average rating of the books by that author.

Yes, I understand that. I hope that I included that in my question, What do you think about the rest of my question?
The whole concept of rating could actually be used to greatly amplify research capability without major rebuilding.
Code for scraping this data is all over the net, it's only 1 to 3 numbers.

[kovidgoyal]

that would be scraping data of a particular kind from a particular source, not something I mainterested in personally, but calibre has a very powerful plugin infrastructure, anyone is welcome to write a plugin to scrape this data.