Classification to organize large data sets is a critical tool in the drive to business value. So why is it that CIO’s desperately want to classify their information, but are not implementing and taking advantage of it?
ABBYY Smart Classifier is a document classification system with
Smart Classifier can classify a large variety of documents and texts in many different languages.
Various specific features of the documents are analyzed and used to assign documents to the categories that were defined by the users.
Smart Classifier’s text-based classification algorithms based on morphology and granular textual features delivers very high classification precision. Semantic classification places an entirely new set of capabilities in the hands of business and IT professionals. Based on Compreno, ABBYY’s innovative natural language processing technology, Smart Classifier can classify documents based on the linguistically derived meaning of words and sentences, including relationships, facts, and events.
Optimizing classification models for specific categories is a grueling process of selecting document characteristics and model parameters. Then you iteratively have to run tests to get good results. ABBYY Smart Classifier does this “dirty work” for you– automatically making the decisions to optimize the algorithms and deliver consistently high classification quality.
Smart Classifier is a new experience, from the UI to the number of documents needed to train the system. No longer will you require thousands of documents to train a category. Smart Classifier reduces that by a factor of 10 or more.
Using a simple REST API, integrating Smart Classifier into workflow, archives, records management systems, email management, data migration or compliance solutions is a snap - efficient, fast and easy.
In the PDFs below you can find a technical overview (4 pages) and business introduction (3 pages) to ABBYY Smart Classifier.
Smart Classifier form a technical perspective: setup, features, architecture, integration, system requirements.