Brendan Marshall completed BIT (Hons) in 2006 on Optimization of Human-Computer Interaction in High-Volume Data Entry and was supervised by StuartMarshall
It involves research into the user interface issues involved in high-volume data entry (such as you might find, for example, in a bank processing handwritten credit card application forms). In essense we are seeking to prototype a generic user interface for such situations which improves the speed and accuracy with which data can be captured. This will include: *Improving the speed at which users can key from an on-screen scanned image of a form (this is known as "heads-up" keying in the industry). *Improving accuracy rates which keying from an image. *Introducing the use of ICR
(Intelligent Character Recognition) into the equation: **Improving the rate at which human users can study and correct ICR errors in forms. **Improving accuracy and detection rates for ICR errors.
Rapid, vertical prototyping is being used for the creation of the prototypes being tested. Each vertical prototype (or major iteration of an existing prototype) will test a particular feature or set of related features.
User testing will be involved in evaluating prototypes, specifically with regard to speed and accuracy rates.
Some of the key technologies being used in this project:
- Java with SWT
- The main programming language and interface technology being used for the application.
- XML (including XSD & XSL-T)
- What fields are there?
- Where should they be placed on the interface?
- Where are they located on the corresponding scanned image?
- What constraints are placed upon them? (e.g. Numeric between 100 and 900)
- What relationships exist between them? (e.g. Field A + Field B <= Field C)
- Kadmos ICR (http://www.rerecognition.com/)
- Used for automatic recognition of handwritten characters