Populations are aging and the prevalence of chronic disease, persisting for many years, is increasing. The most common, non-communicable chronic diseases in developed countries are; cardiovascular disease (CVD), type 2 diabetes, obesity, arthritis and specific cancers. Chronic diseases such as cardiovascular disease, type 2 diabetes and obesity have high prevalence and develop over the course of life due to a number of interrelated factors including genetic predisposition, nutrition and lifestyle. With the development and completion of human genome sequencing, we are able to trace genes responsible for proteins and metabolites that are linked with these diseases. A computerized model focused on organizing knowledge related to genes, nutrition and the three chronic diseases, namely, cardiovascular disease, type 2 diabetes and obesity has been developed for the Ontology-Based Personalized Risk Evaluation for Chronic Disease Project. This model is a Protégé-based ontological representation which has been developed for entering and linking concepts and data for these three chronic diseases. This model facilitates to identify interrelationships between concepts. The ontological representation provides the framework into which information on individual patients, disease symptoms, gene maps, diet and life history can be input, and risks, profiles, and recommendations derived. Personal genome and health data could provide a guide for designing and building a medical health administration system for taking relevant annual medical tests, e.g. gene expression level changes for health surveillance. One method, called transductive neuro-fuzzy inference system with weighted data normalization is used to evaluate personalized risk of chronic disease. This personalized approach has been used for two different chronic diseases, predicting the risk of cardiovascular disease and predicting the risk of type 2 diabetes. For predicting the risk of cardiovascular disease, the National Nutrition Health Survey 97 data from New Zealand population has been used. This data contains clinical, anthropometric and nutritional variables. For predicting risk of type 2 diabetes, data from the Italian population with clinical and genetic variables has been used. It has been discovered that genes responsible for causing type 2 diabetes are different in male and female samples. A framework to integrate the personalized model and the chronic disease ontology is also developed with the aim of providing support for further discovery through the integration of the ontological representation in order to build an expert system in genes of interest and relevant dietary components.