Pooja Casula
  • About
  • Work

Practical RST Tutor


For a course I took called, 'Personalized Online Learning', I worked on a team to design an online tutoring system called 'Practical RST Tutor'. Rhetorical Structure Theory, also known as RST, is a method of annotation that focuses on identifying relations between sentences. Annotation is commonly used to identify patterns and elements in computer generated text, which can then be converted into a format to feed Machine Learning algorithms. Many annotators are unfamiliar with RST and learning the method can be quite dry and laborious. It typically involves mulling over academic papers and websites, which is a generally un-engaging process.  Identifying a need to enhance the RST learning process, we decided to build an engaging online adaptive tutor to help annotators learn this method more efficiently. This tutor relies on CTAT, Cognitive Tutor Authoring Tools, a software designed to help create tutors and track student progress. 

This Cognitive Tutor was presented as part of a package of annotation resources at a conference on Discourse Parsing and Treebanking. To read the published paper, please click "Link to Publication".

Project Year:
2018 

Project Team: 
Kexin Yang
Mingtong Zhang
Pooja Casula 

Project Advisors: 
Vincent Aleven
​Jonathan Sewall
​

Practical RST Tutor Final Report
Link to Publication
Link to Github
  • About
  • Work