Parascript Goes a Step Further in Intelligent Document Processing with FormXtra.AI 8.3

Parascript, which has been delivering high-performing automation for over 25 years and processing more than 100 billion documents annually, released a new version of FormXtra.AI that again expands support for high performance IDP needs, this time to truly unstructured data such as contracts and other text-heavy documents that lack uniformity.

“One of the needs that we’ve heard repeatedly from partners and customers alike is for the ability to automate analysis and extraction of data within unstructured documents such as what is common within lending documents and health records,” said Greg Council, VP of Marketing & Product Management at Parascript.

Recommended AI News: CytoReason and Pfizer Extend Collaboration to Leverage Machine Learning in Drug Development

With the 8.3 release of FormXtra.AI, Parascript adds trainable, grammatical-based text parsing using natural language processing (NLP) techniques to find needed information often buried within paragraphs of text. Since this capability is part of Smart Learning, customers benefit from the same significant reduction in configuration, optimization, and deployment. Here are a few highlights:

FormXtra.AI 8.3 is enhanced with capability to work with unstructured text and handwriting.

What is unstructured text? If you’ve ever seen an insurance policy or the fine print on your credit card agreement, you’ve seen unstructured information. These are the kinds of important information that do not have lend themselves to traditional data capture techniques using locations or keywords that could specify their location on a page or a document. Using new deep machine learning and NLP technologies as the basis, the customer simply marks target data on samples and the system does the rest. NLP algorithms automatically analyze provided examples, interpret the type and the context surrounding the target data, selects the optimal train scheme depending on the location type, value format and other properties of the target fields, and learns how to reliably locate similar data on documents. For users, this means that Parascript significantly expands the types of applications and the range of documents that can be processed using FormXtra.AI.

Recommended AI News: IBM and SAP Strengthen Partnership to Help Clients Move Workloads from SAP Solutions to the Cloud

SignatureXpert.AI is fully integrated in and with FormXtra.AI 8.3.

SignatureXpert.AI is a next generation product based on the most innovative Artificial Intelligence (AI) methods and approaches in image analysis and pattern recognition and builds upon over 15 years of signature location and verification experience. This results in truly transformational, better-than-human precision automation for a variety of signature verification tasks. SignatureXpert.AI works with documents having a broad range of resolution (96dpi – 300dpi) and successfully locates target signatures on any type of document including those that have handwritten data besides the actual signatures.

Other enhancements and improvements include:

• Added automatic page generation for documents with handwritten fields. FormXtra.AI has a few capabilities that turn the arduous tasks of configuring data extraction into something as easy as pressing a button. The automated page generation functionality has now been expanded with the capability of working with handwritten documents. Simply import the page image samples and associated correct values of the target fields on them – the software does the rest.
• Added predefined documents, such as Canadian Invoices, to the Ready-to-Use definitions collection. This improves product usability and user-friendliness since it offers a fine-tuned ready-to-use solution for a broadly used generic document type.
• Improved and enhanced location and recognition capability for various field types – Amount, Initials, Date, Credit Card Number, VIN Number, etc. – so that efficiency and flexibility is improved when working with these fields. These fields are not easy to locate and read because of the variety in the formats, broad context, ambiguity when interpreted, etc. Processing documents containing those fields may be quite challenging because those have various forms of layouts, poor quality, complex background, etc.

Recommended AI News: Siemens Adds Intelligence-Based Design to Xcelerator Portfolio With Latest Release of NX

[To share your insights with us, please write to sghosh@martechseries.com]