Methods for automated essay analysis

1. A processor-implemented method for creating a mathematical model for use in identifying discourse elements, the method comprising:

receiving a plurality of first essays relating to a particular subject;

wherein each first essay is in an electronic format;

receiving annotations for each first essay;

wherein each annotation identifies at least one discourse element;

identifying features with a processor;

wherein each feature is exhibited by at least one identified discourse element;

computing empirical frequencies with a processor;

wherein each empirical frequency relates to the presence of a feature with respect to the identified discourse elements across the plurality of first essays;

associating with a processor each empirical frequency with the related identified discourse element; and

utilizing the empirical frequencies to select discourse elements in at least one second essay.

2. The method of claim 1 , wherein the annotations are prepared by human annotators.

3. The method of claim 1 , wherein the annotations are indicated by the bracketing of the at least one identified discourse elements within starting and ending tags.

4. The method of claim 1 , wherein the features comprise positional features, word choice features, or rhetorical structure theory features.

5. The method of claim 1 , wherein the step of identifying features comprises:

determining positional features;

determining word choice features; and

determining rhetorical structure theory features.

6. The method of claim 5 , wherein the step of determining positional features comprises:

looping through each first essay;

noting the position of the at least one identified discourse element within the essay; and

determining the number of sentences in the essay.

7. The method of claim 5 , wherein the step of determining word choice features comprises:

parsing the plurality of first essays; and

creating a list of words contained within sentences marked by an annotator as being a thesis statement.

8. The method of claim 5 , wherein the step of determining word choice features comprises:

accepting a list of predetermined words of belief, or words of opinion.

9. The method of claim 5 , wherein the step of determining rhetorical structure theory features comprises:

parsing the plurality of first essays to extract pertinent rhetorical structure theory features.

10. The method of claim 1 , wherein the at least one identified discourse element comprises a sentence in a discourse element category.

11. The method of claim 10 , wherein the step of computing empirical probabilities with respect to a sentence in a discourse element category comprises:

computing the following probabilities related to each identified feature:

a prior probability that the sentence is in a discourse element category;

a conditional probability of the sentence having the identified feature given that the sentence is in a discourse element category;

a prior probability that the sentence contains the identified feature;

a conditional probability that the sentence does not have the identified feature given that the sentence is in a discourse element category; and

a prior probability that the sentence does not contain the identified feature.

12. The method of claim 1 , wherein the electronic format is ASCII.

13. The method of claim 1 , wherein the plurality of first essays are received as stored or directly entered electronic text, or are received using a method comprising:

scanning paper forms of the plurality of first essays; and

performing optical character recognition on the scanned paper essays.

14. A system for creating a mathematical model for use in identifying discourse elements, the system comprising:

a data processor; and

computer-readable memory in communication with the data processor encoded with instructions for commanding the data processor to execute steps comprising:

receiving a plurality of first essays relating to a particular subject;

wherein each first essay is in an electronic format;

receiving annotations for each first essay;

wherein each annotation identifies at least one discourse element;

identifying features with a processor;

wherein each feature is exhibited by at least one identified discourse element;

computing empirical frequencies with a processor;

wherein each empirical frequency relates to the presence of a feature with respect to the identified discourse elements across the plurality of first essays;

associating with a processor each empirical frequency with the related identified discourse element; and

utilizing the empirical frequencies to select discourse elements in at least one second essay.

15. The system of claim 14 , wherein the features comprise positional features, word choice features, or rhetorical structure theory features.

16. The system of claim 14 , wherein the step of identifying features comprises:

determining positional features;

determining word choice features; and

determining rhetorical structure theory features.

17. The system of claim 16 , wherein the step of determining positional features comprises:

looping through each first essay;

noting the position of the at least one identified discourse element within the essay; and

determining the number of sentences in the essay.

18. The system of claim 16 , wherein the step of determining word choice features comprises:

parsing the plurality of first essays and creating a list of words contained within sentences marked by an annotator as being a thesis statement; or accepting a list of predetermined words of belief, or words of opinion.

19. The system of claim 16 , wherein the step of determining rhetorical structure theory features comprises:

parsing the plurality of first essays to extract pertinent rhetorical structure theory features.