Method and system for redundant data capture from scanned documents

Abstract

Embodiments of the invention may be used to capture student markings on assessment documents. Embodiments of the invention may be used for selected response questions or constructed response questions. Embodiments of the invention may be particularly suited to process exceptional marks.

1 ) A computerized method to capture data from assessment forms, comprising:
 scanning an assessment form to create scan data;   storing the scan data;   performing a first automated extraction technique on at least a part of the scan data to create a first plurality of responses;   determining a confidence indicator for at least one of the responses in the first plurality of responses;   performing a second automated extraction technique on at least a part of the scan data to create a second plurality of responses;   determining a final plurality of responses based on the confidence indicator and the last plurality of responses created; and   at least one of storing or reporting the final plurality of responses.   
   
   
       2 ) A method, as in  claim 1 , wherein the first automated extraction technique is optical mark recognition. 
   
   
       3 ) A method, as in  claim 1 , wherein the assessment form comprises selected response areas. 
   
   
       4 ) A method, as in  claim 1 , wherein the assessment form comprises constructed response areas. 
   
   
       5 ) A method, as in  claim 1 , wherein the assessment form has pre-printed marks, and at least one of the first or second extraction technique further comprises processing at least a part of the scan data corresponding to the pre-printed marks. 
   
   
       6 ) A method, as in  claim 1 , further comprising identifying at least a part of the scan data that corresponds to a student response. 
   
   
       7 ) A method, as in  claim 1 , wherein at least one response in the first plurality of responses is a derived data element. 
   
   
       8 ) A method, as in  claim 1 , further comprising performing a third automated extraction technique on the scan data to create a third plurality of responses. 
   
   
       9 ) A method, as in  claim 1 , wherein at least one answer from the final plurality of responses is awarded partial credit. 
   
   
       10 ) A computerized system to capture data from assessment forms, comprising:
 a first scanner;   a computer readable medium coupled to the first scanner and operable to store scan data generated by the first scanner scanning an assessment form;   a processor coupled to the computer readable medium;   software, encoded on the computer readable medium, the software including a plurality of instructions, which, when executed by the processor causes the processor to:   perform a first extraction technique on at least a part of the scan data to create a first plurality of responses;   determine a confidence indicator for at least one of the responses in the first plurality of responses;   perform a second extraction technique on at least a part of the scan data to create a second plurality of responses;   determine a final plurality of responses based on the confidence indicator and the last plurality of responses created; and   at least one of store or report the final plurality of responses.   
   
   
       11 ) A system, as in  claim 10 , further comprising a resource allocation system coupled to the processor. 
   
   
       12 ) A system, as in  claim 10 , further comprising software, encoded on a computer readable medium, to decompress image data. 
   
   
       13 ) A system, as in  claim 10 , further comprising a second scanner; wherein the first scanner is at a first test site, and the second scanner is at a second test site. 
   
   
       14 ) A computerized method to conditionally extract data from an assessment form; comprising:
 storing, on a computer readable medium, a plurality of techniques to process an image of an assessment form;   executing a first technique to process the image of the assessment form;   determining an indicator of the accuracy of the first technique with respect to a first answer;   comparing the indicator with a confidence criterion;   conditionally executing a second technique to process the image of the assessment form based on the indicator; and   determining a response based at least in part on the last technique executed.   
   
   
       15 ) A method, as in  claim 14 , further comprising lowering the value of the indicator. 
   
   
       16 ) A method, as in  claim 14 , wherein the confidence criterion is received from an education authority. 
   
   
       17 ) A method, as in  claim 14 , wherein the second technique is selected based the indicator. 
   
   
       18 ) A method, as in  claim 14 , wherein there is only one image of the assessment form. 
   
   
       19 ) A method, as in  claim 14 , further comprising identifying target locations in the image of the assessment form. 
   
   
       20 ) A method, as in  claim 14 , wherein the indicator is based on a darkness level of the first answer. 
   
   
       21 ) A method, as in  claim 14 , wherein the indicator is based on a shape of the first answer. 
   
   
       22 ) A method, as in  claim 14 , wherein at least one image to process the image comprises character recognition. 
   
   
       23 ) A method, as in  claim 14 , wherein comparing the indicator with a confidence criterion further comprises evaluating external data in conjunction with the indicator. 
   
   
       24 ) A method, as in  claim 14 , further comprising identifying a discrepancy. 
   
   
       25 ) A computerized system to conditionally extract data from an assessment form; comprising:
 a computer readable medium;   a processor coupled to the computer readable medium;   software, encoded on the computer readable medium, the software including a plurality of instructions, which, when executed by the processor causes the processor to:   store, on the computer readable medium, a plurality of techniques to process an image of an assessment form;   execute a first technique to process the image of the assessment form;   determine an indicator of the accuracy of the first technique with respect to a first answer;   compare the indicator with a confidence criterion;   conditionally execute a second technique to process the image of the assessment form based on the indicator; and   determine a response based at least in part on the last technique executed.