HUMAN-CENTERED AI

Why would you guess about facts, if human-centered AI can reveal them for you? We offer personalized solutions that help you understand the information stored in structured and unstructured data. We make data searchable, extract meta-data (e.g. named entities, timestamps) enrich data (e.g. classification, topic modelling, sentiment and emotion scores) and visualize the information. Our complex approach applies state-of-the art methods of machine learning and NLP.

About human-centered artificial intelligence

Human-centered AI helps you get empirical insights into your data through the following steps:

  • data gathering (What type? E.g. text, blog or image
    From where? E.g. data harvested with TAS Data Collector or own data)
  • data cleaning, validation
  • annotation
  • transforming, feature engineering
  • feature extraction
  • model training, parameter tuning
  • supervised machine learning: classifier
  • unsupervised machine learning: topic modeling, clustering
  • model evaluation
  • web services
  • visualization (What type? E.g. interactive dashboard, print visualization)

Blog

The latest news, project reports and description of interesting scientific phenomena are available on Precognox Blog.

Use cases

Get to know our projects on migration and on a classification issue.

Learn more about our text analytics solutions on the TAS Text Analytics System product site.