1 / 22

CAPIT: An Intelligent Tutoring System for Capitalisation and Punctuation

CAPIT: An Intelligent Tutoring System for Capitalisation and Punctuation. ICTG Group Department of Computer Science University of Canterbury. Overview. Introduction Punctuation/Capitalisation Domain Constraint-Based Modeling (CBM) CAPIT Evaluation Study Results Conclusion. Introduction.

lou
Download Presentation

CAPIT: An Intelligent Tutoring System for Capitalisation and Punctuation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CAPIT: An Intelligent Tutoring System for Capitalisation and Punctuation ICTG Group Department of Computer Science University of Canterbury

  2. Overview • Introduction • Punctuation/Capitalisation Domain • Constraint-Based Modeling (CBM) • CAPIT • Evaluation Study Results • Conclusion

  3. Introduction • CAPIT is the second constraint-based ITS • Domain is English punctuation and capitalisation for school children • Basic usages of capitals, commas, full-stops, quotation marks

  4. Introduction • First evaluation of CAPIT held in April 2000 • Results indicate that children gradually learned the rules of the domain • Children much more motivated by CAPIT than by traditional pen-and-paper exercises

  5. Punctuation/Capitalisation Domain • Check-and-correct: student checks for errors, if any, and corrects them • Completion exercise: student must punctuate and capitalise an unpunctuated, uncapitalised piece of text • Latter type of exercise chosen

  6. Punctuation/Capitalisation Domain • Example: the teacher said open your books • Student submits: The teacher said, “open your books”. • Two errors!

  7. Punctuation/Capitalisation Domain • open should be capitalised • Period should be inside the quotation • Correct Answer: The teacher said, “Open your books.”

  8. Punctuation/Capitalisation Domain • Another example: theres a bee buzzing past me its taking its honey back to its hive i hope it knows its way home

  9. Constraint-Based Modelling (CBM) • SQL-Tutor is another CBM tutor • Domain knowledge represented by a set of constraints • A constraint is a pattern of form <Cr, Cs> • If a solution matches the Cr then it must also match the Cs, else something is wrong

  10. CAPIT • Designed for 10-11 year old schoolchildren • Interactive system for punctuating and capitalising text • Problems must be designed by a teacher • 45 problems and 25 constraints • Motivation: points and reward animations

  11. CAPIT • Constraints cover: Capitalisation of sentences and names Ending sentences with a full-stop Contraction of is and not Denoting ownership Direct speech etc etc

  12. CAPIT • A problem consists of a list of words • Each word has one or more tags

  13. CAPIT • Example:

  14. CAPIT • A constraint consists of a Cr and a Cs • In CAPIT, each constraint also has a feedback message • A Cs is a set of tags • A Cr is a regular expression

  15. CAPIT • Example: Cr = {NAME-OF-PERSON} Cs = ^[%SYMBOLSET%]*[A-Z0-9] Msg = Each word in a person’s name should start with a capital! • More examples in the paper

  16. Evaluation Study • April 2000 • Westburn School, Christchurch • 28 10-11 year olds working in pairs • Four 30-45 minute sessions over 1 month • Preliminary evaluation for a more comprehensive evaluation that followed in June 2000

  17. Evaluation Study • Averages per student: 89 attempts at 28 problems 30 seconds per attempt 45 minutes interaction time 21 out of 45 solved problems 7 abandoned problems 181 violated constraints, with feedback on 68

  18. Evaluation Study

  19. Conclusion • This version of CAPIT had no long-term student model • Next problem was selected randomly • Most appropriate error message also selected randomly (from set of violated constraints)

  20. Conclusion • Current version of system has Bayesian network student model. • BN built using data acquired during the April evaluation • Subsequent evaluation of that complete system held in June 2000 (see IJAIED paper)

More Related