Skip to content

Netflix Recommendations and Library Book Recommendations

This article made me think.  What can we learn from Netflix about the changes in their business from physical DVD distribution to streaming media distribution?  This is a pretty good metaphor for the transition libraries are facing.

So this article made me think about the following questions:

1. What information are libraries missing in understanding e-reading behaviours in aggregate?

2. Do I think that people read more ‘popular’ stuff than ‘literary’ stuff that a reader survey might say?

3. Is there a way to empower personalized recommendations for library cardholders using the reading data that improves on or supplements the cardholder borrowing records?

4. Is it time for some research into new algorithms for cardholder behaviour? (OCLC)

5. Are there some vendors out there like LibraryThing, Goodreads and Bibliocommons that have explored and shared this?

6. Can some of this apply to database use? (Like Gale Cengage’s work with Foresee)

7.  Has anyone gotten under the hood at Amazon recommendations or B&N, Sony or Kobo?   Personally my hypothesis is that cardholder behaviour is slightly different than consumer retail buying behaviours in many ways but that will change with greater anonymity of e-books (there a reason why e-books helped erotica develop as a class of literature during the e-book era (Can you say 50 Shades of Gray?)

8. With innovative emerging applications like ReaderFirst in libraries, it is a huge opportunity for libraries to better understand reader behaviours and preferences and improve recommendations.



Netflix knows you have horrible taste

Interviews with company executives reveal its recommendation engine is more comprehensive than you might think

“Netflix caters to your secret terrible tastes and is even working on getting the algorithm to spit out movies based on what time of the day or week it is.”



Posted on: August 13, 2013, 7:10 am Category: Uncategorized