This paper explores the potential of Twitter as a source of public health information. The authors introduce the Ailment Topic Aspect Model (ATAM) to analyze health-related tweets and identify various ailments, including allergies, obesity, and insomnia. They extend ATAM by incorporating prior knowledge to improve its accuracy and apply it to tasks such as tracking illnesses over time and location, measuring behavioral risk factors, and analyzing symptoms and treatments. The results show strong quantitative correlations with public health data and qualitative evaluations of model output, suggesting that Twitter has broad applicability for public health research. The study demonstrates that Twitter can provide valuable insights into public health, including syndromic surveillance for influenza, geographic behavioral risk factors, and the analysis of symptoms and medications. The authors compare their results with existing public health data and find that ATAM+ (an improved version of ATAM) performs well in terms of correlation with public health data. They also show that Twitter can be used to track the spread of seasonal allergies and influenza over time and geography. The paper also discusses the limitations of using Twitter for public health research, including the need for more data, the coarse geographic resolution of the data, and the demographic limitations of Twitter users. The authors conclude that while Twitter has significant potential for public health research, there are still challenges to overcome, and more data and advanced techniques are needed to fully realize its potential. The study highlights the importance of using social media data for public health research and suggests that future work should focus on improving the accuracy and scope of such analyses.This paper explores the potential of Twitter as a source of public health information. The authors introduce the Ailment Topic Aspect Model (ATAM) to analyze health-related tweets and identify various ailments, including allergies, obesity, and insomnia. They extend ATAM by incorporating prior knowledge to improve its accuracy and apply it to tasks such as tracking illnesses over time and location, measuring behavioral risk factors, and analyzing symptoms and treatments. The results show strong quantitative correlations with public health data and qualitative evaluations of model output, suggesting that Twitter has broad applicability for public health research. The study demonstrates that Twitter can provide valuable insights into public health, including syndromic surveillance for influenza, geographic behavioral risk factors, and the analysis of symptoms and medications. The authors compare their results with existing public health data and find that ATAM+ (an improved version of ATAM) performs well in terms of correlation with public health data. They also show that Twitter can be used to track the spread of seasonal allergies and influenza over time and geography. The paper also discusses the limitations of using Twitter for public health research, including the need for more data, the coarse geographic resolution of the data, and the demographic limitations of Twitter users. The authors conclude that while Twitter has significant potential for public health research, there are still challenges to overcome, and more data and advanced techniques are needed to fully realize its potential. The study highlights the importance of using social media data for public health research and suggests that future work should focus on improving the accuracy and scope of such analyses.