Diseases are affected and altered by different diets in multiple ways. Although diet is an important factor, there is a lack of reliable information related to disease and diet associations. The associations can only be known by reading biomedical research papers as no such dataset is readily available. Manual extraction of such associations is a time-consuming process, so in this paper, we have developed Disease Diet Associations Curator and Explorer (DIDACE) for automatically curating and further exploring disease-diet association database. A two-phase approach has been followed which includes curation of medical literature in the first phase so as to quantify the strength of association of different diseases and diets. In the second phase, generated database is further analyzed to predict the nature (harmful or helpful) of unknown associations. This is done by performing sentiment analysis and machine learning using curated database. The database, thus generated, comprises both nature and strength of Disease-Diet associations. Such databases might prove to be a useful resource for medical and health informatics researchers for understanding complex interdependencies of different foods and diseases.