Structured and Unstructured Documents: What are the Differences?

First, why does that even matter? You’re probably asking that question because you’ve been doing research on how to make your data entry tasks easier at work. This question is a great one and one of the first steps in understanding the options available to you for automating the processing of your company’s documents; paper, pdf, or otherwise. Below you will see several examples of documents. We provide examples of structured documents, unstructured documents, and even semi-structured documents. 

Structured Documents

Structured documents you might be familiar with in the form of:

• Surveys
• Questionnaires
• Tests
• Claim Forms

Surveys

In the mail you may have received census survey forms that ask you to input your data into structured boxes. The reason for this is that utilizing OCR, scanners are able to recognize the letters and numbers that you write in the boxes much easier and more accurately than free-form text. Of course, there is still room for error with scanning and capture, however, using constrained text boxes and response ‘bubbles’ helps tremendously in interpreting the data provided.

Questionnaires

A questionnaire, much the same as a survey, is also an example of a structured document. The template is set up to accept all responses in the same format from each recipient.

Tests

Remember Scantron forms that you used to record your answers from tests you took in school? Those are also structured documents!

Claim Forms

Ever shopped at Menards and filled out one of those rebate forms to get your $6 back in Menards credit on a purchase you made? How about health insurance claim forms? Or those of you that are road warriors; have you submitted an expense claim form? These are all examples of structured documents as well.

Semi-Structured Documents

Semi-structured documents are also widely used. Examples include:

• Invoices
• Purchase Orders
• Bills of Lading
• Explanation of Benefits
• Invoices

You can probably think of several styles of invoices. For the most part though, they all contain the company name, address, and phone number, invoice and/or purchase order number, due dates, line items, and total amounts due. While they may not all be laid out the same, you can train your OCR software to recognize each of these different formats to scan and capture the critical information off the invoice and automate the process of getting the invoice paid and logged in your ERP.

Purchase Orders

Similar to the invoices, purchase orders typically contain the same information as the invoices do. Not all PO’s look the same, but with machine learning, you are able to train your OCR software to recognize the important data and match it to your master database, thus automating the input and workflow processes.

Bills of Lading

Does your company ship goods? If so, you are likely familiar with bills of lading. With many of these forms, there will be a barcode that will enable the workflow process by scanning the barcode and routing the bill to the proper department. Others will just have unique codes specific to the vendor and customer that your OCR software recognizes and processes within your ERP.

Explanation of Benefits

Health insurance EOB’s are a great example of a semi-structured form. They contain the fixed information like health care provider name, address, and contact info, patient name and info, and likely also a bar code, but they also contain things below the structured section that varies patient to patient.

Unstructured Documents

Unstructured documents are just that; documents that can be free-form and don’t have a set structure but are still able to be scanned, captured, and imported.

Some examples are:

• Contracts
• Letters
• Articles
• Memos

Contracts

Contacts are not all made alike, so these are considered unstructured documents. Each company or contractor will likely have their own format, so it’s a little more challenging to train your data capture software to automate the data capture from these items. With flexible layout technology, information can be extracted in relation to a keyword or regular expression regardless of where the information falls on the page.

Letters

A letter is just what you think it is; a customer has written you a letter to terminate their service or maybe to dispute a transaction, you’ll want to have that content accessible in your ERP. Using OCR technology, you can scan the text and have it imported so it is searchable and accessible by all authorized users in your company. With classification technology, you can also determine the type of letter, whether it be a cancellation letter, inquiry, sales, etc. by keyword indicators that have been pre-programmed to indicate whether the author is mad, happy, etc… How great would it be to eliminate unopened and unrouted attachments or paper lying around with the clock running out?

Articles

Was your CEO mentioned in a news article? With OCR, you can scan the article and create a blog post utilizing the content in the article.

Memos

Much like a letter, internal office memos can also be scanned, archived, and classified.

Video

ABBYY has a really great video that explains the way that you can eliminate the need for manual data entry. Take a look:

So what do you think? What can your company automate?

Contact us today to get a demonstration of the software with your documents to see how Optiform can make your life easier.

Sales: email sales@optiform.com or call 317-822-8150