Most Infectious diseases spread within a short period of time in a given population segment. Therefore, early detection of disease activity and rapid response are required to prevent growing numbers of infected patients and excessive medical expenses. A user’s symptom-related expressions on social media can be used to identify new infectious diseases in real time with high accuracy. However, the naming or defining of a new infectious disease is a longer process that happens over a longer period of time. For example, individuals did not use the term “Zika” on social media when expressing their symptoms before the CDC gave the new virus a name. This work proposes a bottom-up approach that uses social media data to discover new infectious diseases by collecting symptom data without prior information, such as disease names. This research can help biomedical professionals expedite identification to prevent diseases from spreading. This may also translate to lower costs for the healthcare system due to more accurate resource allocation decisions.
Lin, S., Tucker, C. S., & Kumara, S. (2017). An unsupervised machine learning model for discovering latent infectious diseases using social media data. Journal of Biomedical Informatics
For more information, please contact: Conrad Tucker, PhD at email@example.com