Gianfranco's Second 🧠

    Hit Rate

    Created: 2022-12-09 15:12
    #note

    Average reciprocal hit rate(ARHR) = ∑i=1n1rankiUsers\frac{\sum^{n}_{i=1} \frac{1}{rank_i}}{Users} -> more attention to top list

    Cumulative hit rate (cHR) -> we should not recommend items that users will not enjoy.

    Rating hit rate (rHR) -> we should recommend items that users liked

    References

    1. LinkedIn course on RecSys

    Tags

    #recsys