We describe an approach to developing next generation health information portals. This prototype portal was developed to address two complementary goals (1) design and create a site where people can explore potential relationships between selected health-related behaviors, policies, and demographic data (2) explore semantic web technologies and linked data as enabling technologies for next generation health informatics portals. Our multidisciplinary team includes population and behavioral scientists, social network scientists, statisticians, and computer scientists focused on creating innovative proof of concept applications that integrate complex health data in understandable and usable ways. Our semantic-web based framework allowed us to design exemplar community health portal applications, with an initial focus on tobacco-related health data such as smoking prevalence and tobacco policies (taxation and smoking bans). We describe our approach, two semantically-enabled tobacco-related applications, and discuss how this approach can be used in a broad spectrum of community health applications.