No More Paper?...No Way!

P.aper is currently waging a war against its electronic forms rival. For some, the ability to download a 1,000-page product manual onto a PDA (personal digital assistant) is sheer bliss. After all, e-form proponents claim, electronic documents can be accessed quicker than paper documents, they save on physical space, and they can quickly be duplicated. Paper advocates have a few arguments in their defense. First, in vertical markets such as healthcare, many government mandates, such as HIPAA (health insurance portability and accountability act of 1996) do not accept electronic signatures. Second, reading information from a computer screen just isn't as clear as reading a piece of (radiation free) paper. Look at the trends happening in forms processing technology today and decide for yourself if paper documents will go the way of the sword.

Trend 1 - Neural Networking
This first trend involves converting paper documents into electronic documents via a special type of OCR (optical character recognition) technology. Neural networking forms processing is changing the traditional rules of forms processing. Traditionally, templates had to be created, which pinpointed certain areas on a page for the OCR technology to capture. This works well for forms such as 1040 forms, which are standardized and processed in high volumes. "But, other forms such as credit card applications, property tax bills, invoices, W-2s, mortgage forms, and purchase orders, which have many designs for the same form can't be easily molded into a static template," says Reynolds Bish, president and CEO of Captiva Software (San Diego), a forms processing solution provider. "Unlike traditional OCR technology which relies on templates to search and capture text from a scanned page, neural networking has the ability to create templates on the fly." For example, a user who wants to scan invoices can tell the neural network that the batch of documents to be scanned is "invoices." After scanning a few examples, the software will automatically recognize similarly structured documents as invoices. Neural networking can be used with OCR and ICR (intelligent character recognition) technology to locate patterns of words and characters on a document such as "Federal income tax withheld" or "INV No." and can be used to identify the dollar amounts or numbers that follow these fields. In these examples, the dollar amounts on a W-2 or invoice numbers on an invoice could be entered into a database or used for indexing purposes.

Trend 2 - Dynamic Voting
Another trend that has emerged over the last few years is known as dynamic voting. Dynamic voting is the process of using multiple OCR/ICR engines to increase the overall accuracy of character recognition. For instance, one engine may use neural networking technology that can learn and remember patterns or characters over time. Besides neural networking engines vendors may incorporate OEM engines from Mitek, Océ, and ScanSoft to complement one another. "With only one engine you don't know whether to trust the results or not," says Emily Deere, director of product management for Cardiff Software (Vista, CA), a provider of business process automation solutions. "With multiple engines, each performing thousands of algorithms and then voting on the best choice among the various calculations, you have a reliable feedback loop. One engine's weakness is offset by another engine's strength."

Trend 3 - Remote Verification
The third important trend happening within forms processing technology involves off-site or remote document verification. For large enterprises that want scanned documents to be verified at a branch location or, perhaps for companies in high compensation areas (e.g. Silicon Valley or New York) that choose to outsource forms verification, remote verification technology is appealing. "Using remote verification can save enterprises a lot of time and money," says Deere. "Typically, all the scanning is done at a central location and then the scanned documents are accessed and verified through an intranet, using a Web browser. This setup results in a thin client solution whereby remote users don't need to install special software on their work stations."

Trend 4 - PDF With Full Text OCR
Another trend - one that makes electronic documents more appealing - centers on PDF (portable document format) technology. Unlike traditional PDF files, which are usually just static images, PDFs with full text OCR allow users to search for key words, copy sections of text, to index not just document names but down to the paragraph level or even down to the word level. Moreover, PDFs created with full text OCR still maintain the inherent benefit of a static PDF, which is a consistent look on different computer screens and across different printers.

For enterprises that still need to retain some paper documents, forms processing technology can be used to reduce the overall paper trail. This will provide enough labor savings to offset the costs of managing a reduced amount of paper documents. And, thanks to forms processing, what initially appeared to be a battle for the ultimate survivor between paper documents and electronic forms could result in a truce.