Data Mining in Homeopathic Materia Medica

Data-driven research stems from the original idea of homeopathy, which can be transferred to the 21st century with modern statistical concepts, especially techniques of data mining.

In preparing a statistical approach to Materia Medica, abstraction of symptoms is pivotal. The main works of Materia Medica were indexed, creating the requirements for analyzing existing data.

A manifold range of objectives are conceivable for analysis of Materia Medica: e.g., checking the quality of the existing data; assessing the prevalence of symptoms; calculating correlations between symptoms; assessing the discriminating power of symptoms; handling of polar symptoms; analyzing cross-references between medicines; calculating domains for each medicine, such as spheres of action, organs and side localization; building a new repertory from scratch.

As a first step, a comparison between data of Materia Medica, prognostic factor research (PFR) and repertories for six selected repertory rubrics was performed, showing moderately high correlations between Materia Medica and PFR.

Methods of data mining applied to Materia Medica can help to analyze existing data to a maximum extent and contribute to the further development of the homeopathic method, both scientifically and practically.