We find that for five recommendations, the resulting recommendation networks are poorly navigable.
For sake of brevity, we report the results for five and twenty recommendations.
For five recommendations, we find that the recommendation algorithms we investigate considerably limit the discoverability and navigability.
We find that the support for berrypicking, a scenario representing dynamic information search, is also not extensive for five recommendations, but improves for 20 recommendations.
The recommendations generated by association rules result in an SCC of (11%) (MovieLens), (59%) (BookCrossing) and (14%) (IMDb) for five recommendations.
For five recommendations, the success rates for the case of genre-based clusters are comparatively low and range up to (12%).
For two recommendations, discovery is at least partially enabled, in particular for matrix factorization, where for BookCrossing ((53%)) and MovieLens ((40%)), a substantial share of the items is already in a mutually reachable component.
Thus, the questionnaire featured 22 questions that were yes/no or selections from a list, seven questions asking students to rank statements with a Likert scale (65 statements in total), and an open-ended request for two recommendations of how to increase STEM major retention.
Most of the first- and second-choice recommendations in the treatment guidelines for skin infections were congruent with commensal S. aureus antimicrobial resistance patterns in the community, except for two recommendations for penicillin.
Our study showed that most of the first- and second-choice recommendations in the treatment guidelines were congruent with AMR patterns in nasal S. aureus in the community, except for two recommendations for penicillin.
