Abstract: 【Purpose/significance】Making use of user portraits and personalized recommendation algorithms to implement book recom?
mendations in smart libraries.【Method/process】Starting from the three data dimensions of constructing the natural attribute, interest
attribute and social attribute of the user profile of the smart library, the similarity calculation method is used to calculate the similarity
of readers and books in different dimensions, so as to realize the virtual book recommendation based on similar readers and similar
books, and expounds the realization of location-based physical book recommendation with the help of the advanced technology of the
smart library.【Result/conclusion】The experimental results show that using user portraits for smart library book recommendation can
improve the personalized service capabilities of library as well as achieve precise recommendations for readers.【Innovation/limitation】
Book recommendation based on user portraits combines recommendations from multiple dimensions, realizing personalized recommen?
dation of virtual and physical books in smart libraries, improving the quality of recommendations, and having a certain degree of im?
provement in personalized services of smart libraries reference meaning. The limitation is that the selected readers and the number of
books is small.