EXTRACT INFORMATION AND CLASSIFY
TEXT FROM DOCUMENTS
The document or text classification module of Information Discovery allows customers to create Artificial Intelligence applications based on text data. We provide classification and clustering techniques based on advanced Text Mining and Machine learning algorithms. They can be accessed both via a powerful graphical user interface or with a simple web interface.
Document classification can be integrated with any project through a web API. By this, our customers can implement capabilities to facilitate sentiment analysis, content monitoring, technology categorization, predictive coding, clustering, alerting and concept searching.
Machine learning techniques massively support human experts in complex annotation and labeling tasks. The expert, instead of programming a rule for every possible outcome, provides a set of training data that shows examples of how the decision should be made. Computers learn from the experience of information professionals and produce useful predictions on new, unseen examples after being trained on a learning data set.
Automatically categorizing large data sets of documents with a high number of (hierarchical) categories while still opting for excellent prediction quality requires a sufficient number of learning data. The concept of active learning minimizes the effort of manual creation of such data by intelligent data sampling and iterative supervised learning.