Multi-View Clustering of Clinical Documents Based on Conditions and Medical Responses of Patients

Clinical documents from electronic patient record system has rich data which addresses the details about patient's disease, injuries, medication response. Patient's data in the clinical documents are clustered into groups based on conditions (symptoms names) and medical response (medication/drug names). The documents with patients who are affected with disease, not affected, and are in hypothetical condition information are clustered using single-view algorithms like Non-negative matrix factorization (NMF) and K-Means algorithm based on cosine similarity and multi-view algorithms like (Multi-View NMF). Multi-view clustering is used to cluster the documents by finding the relationship between different views. The comparison among the algorithms are made to identify the best clustering method to give priority for the patients who are in hypothetical condition.