Collaboration-based medical knowledge recommendation

摘要: Purpose: Clinicians rely on a large amount of medical knowledge when performing clinical work. In environment, organizations must exploit effective methods seeking and recommending appropriate in order to help clinicians perform their Method: Aiming at supporting search more accurately realistically, this paper proposes collaboration-based recommendation approach. particular, the proposed approach generates clinician trust profile based measure factors implicitly from clinicians' past rating behaviors items. And then generated is incorporated into collaborative filtering techniques improve quality recommendation, solve information-overload problem by suggesting items interest clinicians. Results: Two case studies are conducted Zhejiang Huzhou Central Hospital China. One study about drug hold endocrinology department hospital. The experimental dataset records 16 prescribing tracks six months. This shows proof-of-concept other addresses radiological computed tomography (CT)-scan report recommendation. 30 pieces CT-scan examinational reports cerebral hemorrhage patients collected electronic record systems hospital, evaluated rated 19 radiologists radiology 7 neurology department, respectively. provides some confidence will scale up. Conclusion: results show that performs well clinicians, which indicates feasible practice.