A digitally developed methodology to address the preprocessing phase of capturing data from manually filled feedback forms

Abstract

The main objective of my thesis is to generate a new fine tuned Optical Mark Recognition(OMR) which minimize the error count while processing. There are several OMR tools which are used for manual data processing. But, those are not well fine tuned. And there should be a methodology to find out the errors which returns while processing and the reason for the error out come. The newly developed system is consider the option buttons, checkbox and text area as input fields for the questionnaire. Most of the existing OMR systems working under the same set of rules, so that the intention of this system implementation is to provide a system which can dynamically generated forms with varying set of rules. And the next intention is to increase the reading ability of manually filled questionnaires, since the existing systems rejects the filled forms due to several reasons. So the new software provides a more data extraction percentage that existing tools. Then the new system provides the ability to identify the error occuring scenarios while reading the filled forms. The system marks the respective question with red colored box in order to identify the error for the system user. The implemented system is independent of dedicated scanners, so that the system provides the capability of reducing the cost for the full cycle of taking feedback. With regards to the user interface users can create , read and process dynamic forms. And while processing it stores the data in a database which is stored in the computer.