An Electronic Student Notebook
Thomas W. Doeppner
One of the great strengths of the modern residential university is face-to-face interaction among students and faculty. This strength can be enhanced through the appropriate use of technology. We propose to deploy an electronic student notebook. Based on Microsoft’s recently introduced Tablet PCÔ, such a notebook should simplify the taking of class notes, foster better interaction in lectures, promote more informed discussion, and ease collaborative work. Overall, it should provide students and faculty with a better in-class experience and should enhance out-of-class interactions.
Consider the old-fashioned paper notebook. We’re all familiar with it and need no instruction on its use. We go to class or a meeting and take notes in it. We can annotate any printed matter that is handed out and stuff it in the notebook. Later we can go over our notes and handouts, annotate them further, and perhaps share them with others (with the help of a copying machine).
Notebooks are easy to use and are universal. The only compatibility issue we have to worry about is paper size. Distributing handouts (though not producing them) is uncomplicated. Everything can be annotated using the one and only annotation interface. Notebooks are less subject to failures than computers: what we write in them will probably survive power problems, viruses, and being dropped from moderate heights.
Notebooks have their limitations, however. Sharing one’s documents at a meeting is cumbersome, even with a copier nearby. Loss of a notebook is catastrophic. It is difficult to organize and to search their contents. Storage capacity is limited.
Despite amazing progress in computer technology over the past decades, only now, with the introduction of the Tablet PC, are we seeing the beginnings of a useful electronic notebook. The Tablet PC, a technology from Microsoft that has been licensed to a number of hardware vendors, has dealt well with a number of issues: it provides a great annotation interface that, among other important features, has the look and feel of a real notebook. In particular, it lets one use a stylus to annotate computer documents as easily as writing on a pad of paper. The characters that one is writing can be recognized by software and converted into text. Tablet PCs include wireless transceivers for access to the Internet (and to each other).
Paper notebooks are a useful tool in face-to-face interaction, whether in a large lecture or a small meeting. Over the past few years we’ve been examining various scenarios for such interaction, designed prototype interaction interfaces, implemented some of them, and tested one in the classroom. Our plan is to leverage the major advances of the Tablet PC and produce an electronic notebook that not only is a sophisticated note-taking and annotation tool, but is also a major advance in the use of computer technology for aiding face-to-face interaction. We take advantage of wireless communication for document distribution and sharing.
This electronic notebook should enhance the interactive learning experience. Students will use the notebook in lectures to receive, organize, and annotate both instructor-supplied lecture notes and other course materials, all of which will be distributed electronically. Students will be free to participate in class without having to take detailed notes. The notes they do jot down will be fully integrated with all the other course materials. Revisions to the instructor’s materials are also distributed electronically; student annotations can be fully reintegrated[*]. Instructors can give quick quizzes with either anonymous or attributed answers and see the results immediately, allowing them to adjust their lectures accordingly. (This is not a new idea; for example, see [Lit].) In discussion sections, participants will be able to walk through documents together, quickly looking at related materials and making group and private annotations. Transcripts of what was examined will be made automatically, allowing easy review afterwards. Students can take advantage of collaboration tools to ease the crafting of group-produced documents. For example, participants in a seminar course can easily put together an annotated compendium of all papers discussed.
The educational benefits of the electronic notebook are in two related areas:
· Efficiency. Much of the overhead of maintaining notes and handouts is eliminated with the electronic notebook. Today one might read electronic copies of materials, but if annotation is required, one prints them out and marks up the hard copy. Managing the hard copy becomes cumbersome: one might put them in binders, but then one has numerous bulky items to store and manipulate. Searching one’s notes and annotations is laborious if the medium is paper. Searching electronic versions is quick and easy. Establishing and using links and cross references between paper documents is time-consuming. Doing it electronically is fast.
· Depth of understanding. For students to understand an area they must get involved with it. Doing homework sets and lab exercises (such as writing computing programs), as well as writing papers, are certainly ways to getting students involved with the area. Performing these tasks requires an understanding of the area, which comes from mentally processing the course materials, searching for key items of information, looking at additional sources, and somehow putting it all together. The purpose of the above tasks is to make the student go through this mental-processing activity. By improving efficiency in working with and acquiring the course materials, we give students more time to get more out of the materials.
We are certainly not the first to apply modern technology to improving education. There are many excellent products that address distance learning, course management, computer-based discussions, materials presentation, collaboration, etc. We mention a few such products below and distinguish ours from them.
Discourse [Dis] gives the instructor the means for pushing lesson activities to students and for carefully monitoring their progress. It is not a note-taking or annotation tool.
Blackboard [Bla] is a powerful course management system that provides course content management, communication tools for collaboration, assessment tools, and management utilities for instructors. It does not offer students a means for gathering their own materials or incorporating them with the instructor’s, and also does not provide annotation and note-taking functionality. However, it complements what we do nicely and we’d very much like to integrate the use of our product with those of Blackboard and its competitors.
ConferenceXP Presenter [Con] is, like our work, based on the Tablet PC. With it, while making PowerPoint presentations from a computer using a projector, one can mark up the slides as if one were writing on plastic transparencies with a pen. This is a powerful presentation tool. Since it is built on the Tablet PC, it comes with Microsoft’s annotation and notetaking software, but it does not provide the means for document organization, caching reconciliation, and communication that we provide.
Another excellent presentation tool is Pebbles [Peb]. It allows one to control presentations and annotate the projected image. It uses a wireless PDA and thus fits in one’s hand. However, it is not an annotation and organizing tool.
This project builds upon earlier work in which we built a simple electronic notebook on an HPC/Windows CE platform and used it in a lecture course taught by the PI (see http:www.cs.brown.edu/research/systems/esn). Students used their electronic notebooks to receive, view, and annotate course notes, which were distributed by multicast over a wireless network in the classroom. Primarily because of the lack of development and other tools on the HPC platform, our software was primitive. The materials were authored in PowerPoint (on a PC) but, to overcome numerous incompatibilities and other problems, we wrote our own browser and annotator. The system had weak security and the annotations could be viewed only using our tools on the HPC platform. Nevertheless, most of the students enjoyed using it and found it, though not ready for general use, an interesting and helpful step in the right direction. We have since redesigned and re-implemented the notebook, this time on Windows 2000 laptops, and are now in the midst of another redesign and reimplementation, this time taking advantage of the .Net framework and Office tools. Though we no longer can use a stylus for input, our new systems are far more robust and usable than the old.
Our plan is to move our current design to the Tablet PC and integrate it with the Tablet PC’s annotation facilities. We are completing work on a secure, reliable multicast-based document distribution facility and intend to integrate this into the Tablet PC platform as well. The Tablet PC integration work should be completed well before the start date of this proposal.
Our electronic notebook design has five conceptual pieces: document storage and retrieval, communication and security, document distribution, annotation, reconciliation. We briefly describe each in turn.
How should users visualize the space of documents? One approach might be that there’s a single, world-wide (or enterprise-wide) distributed file system (DFS) (See [Sat] for an application of this technology to the mobile environment.). Users have their home directory as well as the ability to access the directories of others (subject to security controls). Electronic notebooks would be clients of such a DFS, caching recently accessed documents. An issue here is the degree of consistency provided.
We find the DFS approach too complicated. Maintaining a DFS requires resolution of the consistency issue and widespread acceptance of a particular DFS approach. Following the (paper) notebook analogy, people don’t need to refer to documents in other people’s notebooks; the model in everyone’s head is of private copies of documents stored locally.
Instead, we favor an approach in which all documents are private. They reside on secure and safe servers and are cached on notebooks. Any one user’s documents are accessed only by that user and are hence cached only on her or his notebook. One’s notebook probably holds all of one’s documents, and thus the servers function strictly as backup devices.
Documents are shared by distributing copies, just as with paper notebooks. However, the copies are attributed with the name of the source, e.g., by tagging their names, as in ENProposal.doc[from firstname.lastname@example.org]. This approach extends naturally to web pages. Rather than using a URL, one simply “stuffs into the notebook” a copy of the actual web page, thus avoiding such issues as disappearance and change (e.g., getting a copy of yesterday’s 3pm New York Times web page).
Our electronic notebooks communicate with one another primarily with a reliable multicast protocol of our own design, layered on top of UDP/IP. Depending on the situation, communication is either unencrypted and between unauthenticated, anonymous principals, or is encrypted and between authenticated principals. The former is appropriate for a lecture scenario, the latter for meetings in which confidential information is being discussed (for example exam preparation or grading).
An important issue in inter-notebook communication is establishing their Internet addresses. This is normally done with the use of DHCP (dynamic host configuration protocol), in common use in most universities. This works fine as long as there is a local infrastructure to support it. However, a group of people with notebooks should be able to intercommunicate as long as they are in proximity of one another, regardless of the local networking infrastructure. To allow this infrastructure-less mode of operation, we will support ad-hoc networking, which provides a means for notebooks to communicate with one another using the standard protocols, but without DHCP or any other Internet-provided service. A collection of notebooks operating strictly in ad-hoc mode cannot access the Internet, but they can communicate with one another. Of course, when such a notebook moves into an area where there is support for DHCP along with connectivity to the Internet, it can switch to infrastructure mode and gain full Internet access.
To make this work, we have to deal with a number of issues involving Internet address assignment, naming, and routing, but these are issues we fully understand and expect to have solved and implemented by early in the grant period.
Distributing a document should be easy. If I’m in a meeting and wish everyone to have a copy of a document, I’d like to distribute the copies by doing something as easy as pushing a button. Of course, I want to make certain that only the intended audience receives the document. I may be holding my meeting in Grand Central Station where a dozen other current meetings are using the same technology. I can’t simply multicast my document over the wireless network; I must use appropriate technology to make certain that only my group receives the document.
We are developing authenticated reliable multicast techniques in which a communication key is established and then is securely distributed to all members of the desired group and used to encrypt and decrypt multicast documents. Our current approach is based on SSL; we had considered more direct approaches to n-way key distribution, but decided that the intricacies of doing security right are so involved that it is safest to use well tested existing technology, even if not terribly efficient. This work should be completed well before the start date of the proposal.
Despite our reluctance to support a DFS, there are situations in which it would be useful for members of a group to have access to one another’s documents. For example, though group members have exchanged copies of their annotated lecture notes, they haven’t exchanged the transitive closure of all the documents (such as web pages) that are referred to by these notes. Furthermore, we might be in a situation in which all group members’ notebooks can communicate with one another via an ad-hoc wireless network, but they do not have access to the Internet (we’re in Grand Central Station). To provide group access to all the documents and web pages being used, each party could make a relevant subset of their documents available to others via a “distributed proxy cache” arrangement, in which one can consult one’s neighbors’ caches for a copy of an otherwise unavailable document. If available, the copy is retrieved from a neighboring cache and attributed with the name of the source of the copy.
The distributed proxy cache is not currently supported and probably won’t be for a year or so—it’s not in our critical path.
This is one of the critical areas of our project, and we’re delighted that most of what we need is available on the Tablet PC, through the Windows Journal and the Office XP Pack for Tablet PC. Our requirements are that, using a single, consistent interface, we be able to annotate any sort of document, yet still be able to follow hyperlinks contained in the document. Of course, one should be able to annotate the document that’s the target of the link, retaining both the document and its annotations. (We’d also like to be able to annotate voice and video, but such documents are not currently in the scope of our work.) Furthermore, annotations must be separable from the document being annotated, allowing us to merge annotations, reconcile changes, etc.
The area in which we can go far beyond the capabilities of a paper notebook is providing tools for reconciling documents, i.e., combining multiple related documents into an integrated whole. For example, say that in a lecture the professor hands out a modified version of the notes that the students have laboriously annotated. They’d like to be able to apply their old annotations to the new notes. With some fairly severe restrictions on the form of the changes, we can do this automatically. But, for the general case, we should provide tools to find the differences and commonalities among documents and then copy over annotations where appropriate. We have given this area a fair amount of thought, but have yet to develop an implementation. However, much of what we need is currently available in Microsoft Office, so, particularly for Word documents, reconciliation is working.
The electronic notebook is certainly a fun toy. But it’s also a great aid to education, and not just for those (found particularly in CS classes) who are excited by such toys. Our work under this proposal will provide experimental proof of the previous statement, with some additional research and development of the notebook itself.
As mentioned earlier, we expect our notebook to aid education by dramatically improving the efficiency of many student (and faculty) tasks and by aiding and encouraging students to study material at greater depth. To improve efficiency we will address the following areas:
· Collecting materials. We will use our reliable multicasting functionality to allow the instructor (or other presenter) to quickly distribute materials to a class. Though most of us are accustomed to simply making hard copies available at the beginning of class or providing links to soft copies, the ability to quickly and effectively put materials into the “hands” of students will facilitate efficiency in other areas, as explained below.
· Annotating materials. All electronic materials can be annotated using the stylus. This not only includes lecture notes, homework solutions, and handouts, but any additional material, including web pages, that the student finds relevant. Note that annotation with a stylus on a Tablet PC is far easier and more convenient that annotating using a keyboard and mouse. The interface is essentially “pencil and paper” with the double benefit that everything is recorded electronically and that handwritten characters are recognized by software and converted into ASCII text.
· Note taking. Notes, as distinct from annotations, can be easily taken on “blank pages,” again with the stylus. Of course the notes can be annotated just like anything else and, where warranted, students can share notes.
· Organizing materials. As student collect materials from a variety of sources, they shouldn’t have to spend time determining how to organize them and then maintaining and organization. Furthermore, the course instructor has probably already given a fair amount of thought to how things should be organized and should be able to provide some guidance to the students. We are experimenting with the notion of coursebooks. Using the course web page as the basis, all course material available in electronic form is assembled over the course of the term into what is essentially an electronic book. However, rather than merely being a web page available to all students in the course, which may disappear at the end of the term, students have their private copies along with materials they’ve collected or produced themselves along with their personal annotations. The coursebook evolves over the course of the term as, for example, the instructor provides material to all, such as additional lecture notes, homework assignments, etc., and students add their own materials and annotations, by simply creating links, which automatically create a private copy of web pages, and cross links.
· Searching. An important aspect of organizing materials is being able to find things. While the coursebook notion provides an organization that will probably reflect the order (hierarchical and temporal) in which material is presented, it’s often necessary find things in the course material even when the organization provides little help. Since everything, including handwritten notes and annotations, is in electronic form (and handwritten notes have been converted to text), searching becomes very easy, particularly with modern search tools.
Though improvement in efficiency is important, real improvement in education will come only if students can and do take advantage of this efficiency to obtain a deeper understanding of whatever concepts the course is teaching. My expectation is that by providing the tools that make it easy to explore additional material and relate it other course material, and in particular to annotate this material and provide cross references, students will actually do this and benefit from it. I give examples in the following scenarios. Though I expect the electronic notebook to be of great benefit in a fair number of teaching and learning styles, I address only those with which I’ve had experience as a teacher.
I teach a course each spring on computer networks. I use PowerPoint slides along with notes, which I make available to students before class. Though I try to have the materials ready the night before class so that students can print them out if desired, I sometimes don’t get things ready until just before class. Though I supply notes along with the slides, I expect students to annotate the slides and notes, partly because I might say things that aren’t in the notes (for example, in answer to a question), and because this helps keep students’ attention. However, if I’m late with the notes, students can’t print them[†]. Thus students annotate their paper copies of my notes.
Often in my lectures I would like to present various web pages to class, particularly those that have detailed specifications of some of the communication protocols we’re covering. Though it’s easy for me to simply go to a web page (and display it using the computer projector), students cannot easily both follow what is going on and take notes. Even if they have computers with them, it’s not easy for everyone to get the URL and then go to the web page. Once there, they can’t annotate the web page itself (there have been commercial products that let one do this, but there are none that I know of that are currently available, and those that were relied on a central server that’s not always available). If the class is large or the web server or perhaps proxy server is busy, it might take a fair amount of time for everyone to get to the web page.
Rather than make students go through all this, I copy portions of the web page into my slides and provide a link to the source. Thus students can visit the source on their own after class, but they still cannot annotate it. They can annotate what I copied to the slides and notes, but not if it’s a web page that I visited in response to a question and thus did not have it in my slides and notes.
I cover some protocols fairly briefly, but I expect students to find out more about them on their own. Most of them are described by somewhat lengthy documents on the web. They’re not too bad to read in softcopy form, particularly when there might be only bits and pieces of the document which should be read. But if students want to study these bits and pieces carefully and thus annotate them, perhaps with references to other course material (or make references from other course material), they must print out the entire document (maybe 100 pages), and then work with a rather cumbersome pile of paper. Even the most enthusiastic students tire of this after the fourth or fifth protocol.
Now imagine my course with the electronic student notebook. The course has a web page, of which students have copies as part of their coursebooks. As changes are made to the web page, students retrieve the changes, which are automatically merged into their copies. If lecture notes are ready before class, students can peruse them ahead of time. Otherwise, when students get to class, the notes are immediately available.
As I lecture, students annotate the softcopy of the notes. I display the associated slides on the projector and perhaps make my own annotations to emphasize points or make corrections. If deemed necessary, these annotations are immediately provided to the students as annotations to their copies.
Either as a planned part of the lecture or in response to questions, I might visit a web page or look at some other material. Students immediately get a copy of what I’m displaying and can annotate it. Rather than run the risk of long delays as all students are downloading the page, it is quickly multicast from my notebook to everyone in the class. Furthermore, if the Internet turns out to be inaccessible during my class, or key servers are down, copies of those materials that are cached on my computer can still be distributed to the class, even though they nominally reside on some server out on the Internet.
After class students study some materials in greater depth. Either from links given in class, or from their own searches, they pull in materials to their coursebooks. With the electronic notebook, there’s no need to print the materials. Students can annotate them and set up cross links without adding to the physical bulk of what they’re maintaining for the course. Furthermore students have powerful search capabilities that simply don’t exist on paper-based materials.
The coursebook is thus a personal, dynamic document maintained throughout the course. Though it can be implemented without the electronic notebook, the communication and annotation capabilities of the notebook make it easy to use and thus ensure that it will be used.
During their interactions, both with me and among just the three of them, a lot of papers were discussed, not all of which had been read by everyone. Copies of the papers were exchanged, annotated, and a fair mass of material was accumulated by all. As is common with any sort of scholarly endeavor, a fair amount of time was spent in simply keeping the materials organized and in searching for relevant information from papers and notes already examined.
The electronic student notebook would greatly simplify this task, allowing students to process more material. In meetings we could easily exchange documents electronically, immediately inserting them into the individual coursebooks where they are organized and can be annotated and searched.
After early experiments with Handheld PCs running Windows CEÔ, we have acquired a number of Tablet PCs and have a prototype Electronic Notebook ready for use. We can distribute and annotate both Word and PowerPoint documents, with HTML and PDF next on our list to support. Both private and shared annotations are supported. Group communication is done using a reliable multicast protocol we’ve designed, and we support group authentication and communication privacy.
The systems will undergo testing and further development this summer (2003) and we plan to try them out in Brown classes this fall.
[*] We can do such reintegration now using the capabilities of Microsoft Word and are working on PowerPoint, HTML, and other document types.
[†] Clearly a moral here is that I shouldn’t be late with the notes. I should get them done in time. Better yet, from the students’ point of view, I should have hard copies made for them. I strive to get the notes ready on time and normally do. In the past I have printed the notes for the students, but I was distressed by the number of copies that weren’t picked up and that had to be recycled. Having students print the notes themselves on demand saves trees.