Forms provide great information for companies and organizations - information that can be used to expand the capabilities of CRM in the future, marketing and web initiatives. Every year, companies spend millions of dollars manually converting paper form information into electronic data. This process is time consuming and expensive, and the data entered is often inaccurate.
Today, more and more companies are turning to data capture technologies to automatically extract data from forms and reduce the time, costs and errors associated with manual processing. Recognition and data collection providers now offer advanced technologies that can easily and accurately recognize various fields in a form, including machine print, hand print and cursive text, as well as tags, flags and bar codes.
Automatic data collection is achieved using specialized recognition technologies such as handwriting recognition (Intelligent Character Recognition or ICR), machine typing recognition (OCR or text recognition), unlimited handprint and italic recognition (natural handwriting recognition or NHR) and optical sign recognition (OMR).
Forms design is essential for successful recognition processing. Forms should be well designed to reduce illegible, inaccurate, misinterpreted or missing data. Using special design methods, you can make “forms smarter” and improve the quantity and quality of data that is filled in and automatically recognized.
Getting answers to fill out all the form information
The importance of field accommodation
The placement of key fields, such as name, address, social security number, account number and bar code, is crucial to the data collection process. Key fields should be located closer to the top of the page, when possible, to account for any shifts during the scan. If important information is placed at the top, the offset is less than at the bottom of the page, which reduces the likelihood of processing errors.
Use captions
Signatures tell the respondent what information should be filled in within a specific area. If possible, the header should be placed outside the box — above or to the left — in order to eliminate the noise and better invoke information that is recognized. It identifies the information that needs to be filled out and helps guide the reader. Signatures should NOT be used under a line or field where they can be mistaken for the heading below it.
Preprinting special characters for specific fields can also improve data quality. If left at the discretion of the respondent, the information cannot be filled out in the required format. For example, respondents often completely exclude a region code when filling in telephone numbers. When requesting this information, fill in the parentheses and dashes to separate the area code and the seven-digit number.
Partitions — for example, slashes, characters, text fields, or any combination — can be applied to some of the most common fields, including name (first, middle initial, last), phone number, date, social security number, etc. show the correct format, especially for special fields and characters, improves the quality of the information that completes and helps ensure automatic recognition.
Form design for maximum recognition accuracy
Properly designed forms allow better image processing, so image quality becomes the only problem. The following guidelines will help form managers improve their design for optimal performance.
Tips for "key" fields handled by OCR
The more intuitive the form, the more likely it will be executed accurately. The “Key” fields must conform to standardized formatting and placement, including:
o Name - the name and surname must be identified using two separate fields, the name of which is indicated to the last. The natural tendency is to write in the order of the first, then the last; any other placement may lead to errors.
o Addresses. Advanced Recognition Technology often includes a USPS Postal Service (USPS) database for checking address information to improve data accuracy and quality. The best way to simplify address recognition is to format information in accordance with USPS standards:
§ Street name, mailbox or Rural Route number
§ Apartment number or number (can be on the same or a separate line as the street)
§ Postal name (city), state and ZIP
Use lines and drawers for improved clarity.
Pre-sealed lines and drawers can be used for many purposes — from paths to guide the respondent through the form to the boundaries, offering breaks between different sections.
Boxes define a field area better than a simple line or open space. Providing answers to a specific area you can work with allows them to adjust their answers to fit into the area, and makes it easier to identify and capture information.
Borders should consist of two to three characters around each side or edge. If possible, additional distance in the field is recommended to take into account deviations when scanning the form.
Medium thickness lines and boxes should be used; broken or false lines may be mistakenly taken as part of the completed information, and heavy lines may block the completed information. Customize the answers to use a blue or black pen to avoid lightly written strokes.
Preventing “noise and intrusion” into forms
An important aspect of design consideration is the amount of noise or intrusion around the fill area. Noise can be caused by dark background colors, heavy graphics and a dense space between fields. For example, on a business respondent card, marketers may prefer projects that include colorful backgrounds and images for contacting correspondents, but these projects may interfere with the process of filling out a form, as well as recognize the completed information. You can consider both needs using lighter colors for the background and applying graphics to the outer perimeters.
Other form attributes
Using special characters to separate fields
Information that is collected frequently can be written in several characters: numeric, alphabetic, and alphanumeric. Numeric fields have the highest read speeds, because there is a smaller set of possible answers (0-9). Alphabetic fields also have high accuracy with a well-defined base of letters. The difficulty is with alpha-numeric fields, where many characters can be a letter or a number. For example, without the use of context, it is difficult to decipher if “1” is the letter “L” in lower case or the number “1”. You can eliminate this confusion by breaking numerical and alphabetic characters where possible. If the field contains both alpha and numeric parts, such as “AA - ####”, then create two separate fields with a dash between them and offer a sample format. Another option is to create six separate boxes, one for each character. In any case, you have reduced the likelihood of the information being filled out incorrectly or incorrectly read by the software.
The role of color in reducing interference
Waste ink provides significant benefits when using OCR, ICR and NHR engines. When people fill out forms, they do not always stay within the lines provided. If ink is not used in a pre-printed form, the data that is included with the respondent may intersect with the pre-printed lines, creating false answers. Drop-down inks can help guide respondents, as well as minimize intrusion between pre-printed information and data completed by the respondent. This preprocessing method allows recognition mechanisms to see only the data that has been entered.
In addition, black texts are not recommended for pre-printed text. The two color options that benefit the respondent and the recognition process are green ink on white paper and red ink on white paper. Both are easy on the eyes and can be easily separated.
Conclusion
When more companies turn to data capture technologies to automatically extract data from forms and reduce time, costs and errors associated with manual processing, the development of intelligent forms has become critical. Choosing the right recognition technology is also an integral part of the data collection process. Recognition technologies, such as Parascript Total Recognition technology, easily capture information about forms, including machine-printed, hand-printed and cursive text, labels, rectangles, and bar codes. Properly designed forms with sophisticated recognition software will improve image processing and reduce the appearance of illegible, inaccurate, misinterpreted or missing data. These technologies will help you take full advantage of the data you collect, giving your organization greater efficiency, cost savings, and customer knowledge to gain a competitive advantage.
