Most capture systems get slower and less accurate as more document types are trained. This problem is especially difficult to manage in cases such as mortgage automation where several hundred document types are involved.
The only way to manage this problem has been through hundreds of hours of fine tuning, analyzing results, adding more examples, retraining and repeating the process again. Every time a new document type is added, the entire trained system needs to be updated.
However, Parascript Document Classification software with its proprietary Cascade Classifier reduces all of this work to the push of a button. Cascade Classification automatically analyzes results, adjusts sample sets, and modifies weights. The result is the industry’s highest level of accuracy and throughput without the typical costs.