Now-a-days Digital information in increased drastically, which is collected and stored from different online system. One of such data is available for patient related data. Generally, the major problem is long-sufferings to be aware their personal record and to illustrate satisfactory terminations. In this work, provide patients with personalized treatments for their ailment/diseases by providing them with a series of recommendation and procedures (other than what they are already using) to avoid the death rate or serious causes and quicker recovery. We are collecting patient profiles and their symptoms for particular diseases. Our proposed recommender systems are using by long-sufferings with supplementary data, helping to better realize their condition of health as characterizing by medical record. In this paper we implement collaborative filtering recommender systems for personalized medicine. We address three issues (1) clustering of patient data (2) to analyze the cosine similarity and Pearson coefficient for collaborative filtering (3) providing better personalized medicine recommender system to individual patients. The main objective of our proposed approach is to recognize long-sufferings who are comparable disease symptoms and originate insights from the medical records of analogous patients to give personalized predictions.