SideEffectPTM: an unsupervised topic model to mine adverse drug reactions from health forums.

ABSTRACTAutomatic discovery of medical knowledge using data mining has great potential benefit in improving population health and reducing healthcare cost. Discovering adverse drug reaction (ADR) is especially important because of the significant morbidity of ADRs to patients. Recently, more and more patients describe the ADRs they experienced and seek for help through online health forums, creating great opportunities for these forums to discover previously unknown ADRs. In this paper, we propose a novel unsupervised approach to tap into the increasingly available health forums to mine the side effect symptoms of drugs mentioned by forum users. Our approach is based on a novel probabilistic mixture model of symptoms, where the side effect symptoms and disease symptoms are explicitly modeled with two separate component models, and discovery of side effect symptoms can be achieved in an unsupervised way through fitting the mixture model to the forum data. Extensive experiments on online health forums demonstrate that our proposed model is effective for discovering the reported ADRs on forums in a completely unsupervised way. The mined knowledge using our model is directly useful for increasing our understanding of more challenging ADRs, such as long-term side effects, drug-drug interactions, and rare side effects. Since our approach is unsupervised, it can be applied to mining large amounts of growing forum data to discover new knowledge about ADRs, helping many patients become aware of possible ADRs.