Retro school children University of Washington Computer Science & Engineering
 Student Projects
  CSE Home   About Us    Search    Contact Info 

Student Projects

Last update 12/19/2004

Here are a collection of possible student projects. There is a significant range of projects here - undergraduate projects, quals projects, and research projects. There is possible RA funding associated with the Classroom Presenter Project. If any look interesting, please get in touch with me. I will be adding to this list (and maybe even removing some projects as they get completed!). I also have interests in supervising projects related to the Tablet PC, in Computer Science Education, and in International Education.

The Classroom Presenter Project

My research is currently focused on the Classroom Presenter Project. Classroom Presenter is a Tablet PC based presentation system where the instructor writes on electronic slides with digital ink. The system is starting to get fairly broad use - I estimate that about 200 courses have used Classroom Presenter, so we are starting to get some interesting data on real usage. Classroom Presenter was developed as a distributed application which allows interaction with student devices. This opens up some very interesting scenarios for exploring support of active learning using wireless devices in the classroom.

The Classroom Presenter system is now mature, which opens up a wide range of research opportunities, some of which are described below.

Digital Ink Based Communication

Ink is a very effective communication tool when used in conjunction with speech and prepared content. Our work has concentrated on the tablet pc, although one could look at other form factors as well. There are many open questions in how people use ink in communication, and how best to develop inking technology to support communication. One important factor is that when communicating with ink and speech, the writing device is likely to recieve only limited attention, so the writing application must designed with a low cognitative load. This would be an interesting domain to explore novel pen based interfaces. There are also many interesting questions related to rendering ink in support of presentation.

Analysis of Lecture Artifacts

It is straight forward to capture and playback digital ink when it is used in presentation. Automatic analysis of speech and ink from lectures could be used to improve classroom applications, as well as to create summaries and indices of recorded lectures. We had an ACM Multimedia paper last year which explored some of the opportunities for analyzing recorded lectures. The analysis problems range from low level problems (such as classify ink between attention, textual, and diagrammatic and match attentional ink to content) to high level problems (identify key frames of a diagram used to make several points, simplify a diagram to present the ink useful for note taking). One very interesting challenge is to combine audio analysis with ink analysis - for example, can speech recognition and handwriting recognition be used together to improve recognition accuracy.

Ink Recognition

There are a number of specific ink recognition problems which come out of this domain. First of all, the general challenge of recognizing "instructor scrawl". Domain specific diagram recognition arises in a number of areas. One I am particularly interested in is recognizing data structure diagrams, such as trees or arrays. An application of this would be in analyzing student submissions (see below). Another very challenging domain for recognition is mathematical handwriting, possibly done in conjunction with audio information.

Classroom Technology (Systems issues)

There are many technology issues in developing a distributed system for classroom use. There are issues with using the devices in a wireless environment and some usage models could run into bandwidth limitations. The Tablet PC technology is developing, and the newest version of the SDK (1.7) provides a lower level access to manipulate ink. This creates opportunities for finer control and higher quality distributed ink.

Ink based pedagogy

Writing has always been a very import tool used in the delivery of lectures. We are interested in understanding how digital ink can be used most effectively in lecture presentation, and how practices vary across disciplines. Recorded lectures give us a very rich source of data to study, and growing use outside of CS (in fields such as Chemistry), allow us to begin to compare across disciplines. Understanding how best to use ink can be tied back into curricular design, as well as to the development of better presentation tools.

Support for active learning

There is a lot of interest in integrating a presentation system with student devices. Our particular interest is in using the integrated system to support active learning, where students use their devices for exercises and then the results of the exercises are fed back into the lecture. One simple scenario we have been experimenting with is "student submissions" where students write on slides, send them back to the instructor, and then the instructor selects from these for display to the class. This engages the student in an activity, and provides the instructor feedback on student understanding and misconceptions. This fall I gave a lecture in Martin Tompa's data structures class to students using Tablet PCs. I felt the lecture was successful, although the experience did raise quite a few HCI issues. We have a large enough collection of Tablet PCs to continue experiments in this direction. Two research questions raised by the student submission scenario are mechanisms to allow the instructor to browse the student results, and automatic analysis of the ink solutions. Another research direction for supporting active learning is to develop the system to support more analysis and integration with the lecture slides.

Evaluation of learning outcomes

The grand challenge for educational technology is to show that the technology improves learning outcomes. One experiment would be parallel lectures with different uses of technology and pre and post tests to evaluate learning outcomes. There is interest in the college of education in participating in this type of work.

Curriculum development

I am hoping to teach a course at some point where students are using Tablet PCs in the class to support active learning. It will be necessary to develop curricular materials to support this. I have found that designing slides specifically for ink use substantially changes the lecture. Designing for active learning is even a bigger step. My tentative plan is to start developing materials around a data structures and algorithms course, anticipating that they will be of use some time in the future.

Classroom Presenter 3.0

I am seriously thinking of having Classroom Presenter rewritten from scratch (even though rewriting working systems is listed as a classic software engineering mistake). There are a number of areas I would like to explore that would require radical changes to the system (such as providing a transparent overlay to allow integration with a wide range of applications). After three years of development, the system is getting progressively more challenging to extend, and there are quite a places where components should be redesigned. For Classroom Presenter 3.0, we will take what we have learned about integrating digital ink and electronic slides in the classroom, and develop a new platform for exploring technology in the classroom.

Specific projects

Here is a list of specific projects tied to Classroom Presenter. (More will be added soon)

CSE logo Computer Science & Engineering
University of Washington
Box 352350
Seattle, WA  98195-2350
(206) 543-1695 voice, (206) 543-2969 FAX
[comments to anderson]