Using AI to Build a Holistic Physchological Profiling Engine


Today’s market is flooded with Natural Language Processing (NLP) tools that allow for an easy extraction of sentiment information from raw text. Whereas these could be a good start to explore the kind of richness that sentiment analysis can bring to the table, much more is needed in order to do this at a level where real actionable insights come out of the raw data. Symanto has been building the most cutting edge AI technology for psychological profiling. We want to present at SEMANTiCS 2018 how we have harnessed deep learning technology to build an NLP engine capable of:
Automatic discovery of topics
Aspect-based sentiment
Emotion classification
Author profiling using six psychological traits
Although our models are all trained on English data, our current technology allows to transform and deploy them on any other language without any human intervention losing only 5% of precision. These models have been tested on real world use cases and our customers used these tool to rebrand themselves, tailor their marketing messages and/or bring about new product designs thanks to a better understanding of their customers. 


Interested in this talk?

Register for SEMANTiCS conference