To the human eye every person has a unique physiognomy. Cutting edge  research in the field of AI-driven Psychometrics revealed the existence of reoccurring patterns related to personality traits below the threshold of human perception.

smile to vote® fractionises facial physiognomy into tens of thousands of tiny, digital components and elements (“features”). Data rooms and data clouds arise from these components and develop patterns, which smile to vote® compares to patterns and structures of substantiated reference data. This technology is based on Facial Scanning, Data Mining, Deep Learning, and Psychometrics.

smile to vote® is a technology, which – with the help of Artificial Intelligence – identifies patterns in physiognomy and derives psychometric and behavioral features from them. In addition to standard Facial Recognition procedures, specific physiognomy patterns are being captured in this process. With the help of these patterns and on the basis of high quality reference data sets, objective predictive models are trained. These predictions are then processed by way of aggregation. They provide high, measurable added value in various electoral areas.

The scientific foundation of smile to vote® is the core of the technology: Numerous validation studies (internal and external) secure the process. Our Scientific Advisory Board actively supports and promotes the development of the technology.

smile to vote® groups physiognomy samples with different characteristics, e.g. all persons showing certain patterns or certain types of political behavior. In this way it is detected which physiognomy patterns are typical for certain groups. Subsequently, typical facial patterns for different groups are detected. This is followed by the conduction of high-dimensional similarity comparisons and predictions about which of these groups a newly characterized, unknown person resembles the most.

The reference data sets consist of millions of individual analysis and contain a number of external criteria (e.g. KPIs) and additional information. The development core of smile to vote® uses Machine Learning and Data Mining algorithms in order to detect connections in the selected data. These patterns serve the modelling modules as a basis for analyses and predictions.