E-nhance lectures: Naber and Köhle: JoDI

E-nhance lectures

Larissa Naber and Monika Köhle
Institute for Software Engineering and Interactive Systems, Vienna University of Technology,
Favoritenstr. 11/188, A-1040 Vienna, Austria
Email: larissa@naber.at

Abstract

Ever more lecturers find themselves forced to Web-enhance their courses out of economic pressure or prestige. Universities trapped between rising student numbers and decreasing budgets are turning to e-learning as the one-stop solution, with little concern for student or teacher needs. An e-(nhanced) learning environment can only be successful if it fulfils students' and lecturers' needs alike. The student needs to be supported in various stages of learning, whereas the lecturer cannot afford to spend more time on generating lecture support materials. Investigation of lecturers' and students' requirements resulted in the concept and design of e-ULE (e-Usable Learning Environment), a university-level teaching and learning environment with a strong focus on usability. To ensure learning materials are helpful for students in any learning situation, from gaining an overview to providing reference, an equally usable authoring tool is required: e-ULE's authoring system is geared towards a typical lecturer, without requiring an undue amount of IT or pedagogical skills, but offers support for academic workflow by supporting tasks like literature research and integration, and collaborative editing in large groups (e.g. with students). Following a usability engineering approach, all features of the e-ULE learning environment are derived from user requirements and usability tests. The main parts of the environment are currently at "proof of concept" stage. The system is open source and relies on several prominent open source projects.

1 Introduction

Despite the technological advances of the the 20th and 21st centuries (Xerox machines, Internet, video, ...) much of our university system is stuck in the middle ages. Large amounts of lecture time is given up to the the act of information transfer. While it is true that a well designed lecture is beneficial to the students' progress, many students claim their lecturer lacks rhetorical and didactic skills, rendering this aspect of lectures virtually useless. Moreover, working or handicapped students might find it difficult to attend lectures, and students unable to learn by listening often find it cumbersome.

A recent study [16] found that students want more "quality time" with their teachers and support in answering specific questions, or exploration of advanced topics. The teacher should not be replaced by an e-learning environment, but by e-nhancing the lecture by shifting the main burden of information transfer to other, digital, media, lecture time is freed for discussions and question and answer sessions. E-nhanced lectures, however, can only be successful if they offer students more than conventional lectures without additionally burdening the lecturer.

The creation of e-courses or Web-based lectures is time-consuming and requires considerable skills on the part of the creator. When looking at major players in the university field (MIT is putting all of its lectures online, see http://ocw.mit.edu/index.html) it is apparent this is a multi-million euro/dollar effort, involving thousands of professionals.

So, should e-learning be left to the big players, or can small universities or even individual lecturers e-nhance their lectures, without driving themselves mad or swamping their students in well-meant but useless e-materials? (First-hand accounts of the problems of introducing e-learning can be found in [9], [10] and [25].)

Usable tools for the average lecturer - not computer scientists, not educationalists, but sociologists, architects, theologists, etc. - are needed to e-nhance lectures.

The remainder of this paper is organised as follows: Section 2 clarifies deficiencies of existing tools. Results of a survey of student requirements are summarized in section 3, and lecturer needs are described in section 4. Section 5 presents the workflow of the academic authoring process, leading to the design concepts of e-ULE, which are presented in section 6 as a system overview and in section 7 as the authoring tool for e-ULE.

2 Why do Existing Tools Fail?

Although there are a number of tools available to e-nhance lectures, many of them are not suited for the academic environment. A detailed comparison of Intelligent Tutoring System tools with e-ULE is found in section 6.2. Section 7 compares approaches of existing authoring tools. Common failings include:

  • Wrong focus: many tools are geared towards the creation of business training, lacking support for data types commonly found in the academic environment (e.g. formulas, quotations, etc.)
  • Skills: although there are excellent tools to build learning materials (complete with animations, ...), these require considerable skill on part of the author.
  • Time: adding training time to the time spent producing the materials, the use of some tools takes too much of the lecturer's time.
  • Money: purchasing tools and hiring a group of specialists to help lecturers with the tools over-stretches the budget.
  • Benefits: some of the easier tools to use do not improve the students' situation - downloading the lecture notes instead of buying them is not much of an improvement.
  • Approach: existing tools fail to consider users' needs properly - usable e-learning must lend itself to support different levels of computer literacy among lecturers and students. Aside from literacy aspects, user needs and requirements have to to be clarified.

3 Student Requirements

In individual interviews that were part of our survey of student requirements [14],[16], we found that - aside from the need for more direct student-teacher communication (compare [21]) - the students' use of learning material changes according to different "learning phases".

  • Phase 1 - reconnoitering: Before deciding to take a lecture or exam students want a detailed overview of the topics covered. Some students refer to is as the "90 minutes notes'' or "executive summary''. Many lectures fail to provide support for this stage.
  • Phase 2 - learning the basics: After deciding to take the exam most students read through the material available (and try to solve the exercises). In this stage they try to focus on the main points. Many lecture notes do not differentiate between essential, important, and additional information. Support is generally there, but it could be improved from the student's point of view.
  • Phase 3 - exam preparation: Most students make custom notes of the topics they need to study further. Most classic materials provide no support for this stage; all excerption is done by hand. Although these excerpts can be of great pedagogical value, they cannot be integrated with original materials efficiently.
  • Phase 4 - reference: Lecture notes should not be worthless after the exam. Human memory being as it is, it is often necessary to return to previous "haunts" to look up what has been forgotten. While some books come with a decent index, most lecture notes do without, so information retrieval can be tedious.
  • Phase 5 - selective retrieval: This phase is set apart from the other four phases: without detailed knowledge of the information provided the student wishes to retrieve selective parts that might be required for a different subject matter. Chances of extracting just the relevant information are slim. Prerequisites for later chapters tend to be scattered across various chapters, forcing the student to canvas the material completely.

To provide added value to the students, a learning environment should take the different learning phases into account. Table 1 lists the stages and the student tasks involved plus the support offered by lectures, printed lecture notes and e-ULE notes.

Other "must have'' features for online materials include print-version (for use while in transit) and offline versions as well as an "intelligent" search function.


Table 1. Learning stages and supporting mechanism: lectures, classical lecture notes and e-ULE lecture notes
  Tasks Lecture Print notes e-ULE
Stage 1 Getting an overview Preliminary discussion TOC and  introduction Mini notes automatically generated from topic summaries
Stage 2 Learning the basics Use exercises to ensure basic concepts are understood before proceeding to more difficult ones Tag basic content Tag basic content
Stage 3 Learning -- Provide structured information Custom views
  Deepen understanding Answer questions References Online discussions, FAQ, integration of new material, annotations, involvement in creation process
  Repeating -- Compile personal versions of the material on paper Use digital system to annotate and modify material, create condensed versions, discuss problems with other students
  Self examination -- Problems and solutions Problems and  solutions
Stage 4 Reference -- TOC, index TOC, index, updates, full text search
Stage 5 Selective retrieval -- Possible, but often cumbersome Automatic retrieval of additional information necessary for understanding the selected topic

Providing students with more efficient, e-nhanced lecture notes requires time on the part of the lecturer, time which often cannot easily be spared from giving lectures, supervising students, battling administration and doing research.

As detailed in section 4, the number of tools involved in the creation of a new course alone can be an impediment. Fewer, easier-to-use tools, especially those designed to support the lecturer (as opposed to support secretaries in writing business letters), will free-up resources. These tools should not only support the authoring process but the whole workflow from start to finish.

Another possibility for cutting down development effort is to integrate students in the creation process. Participating in the creation of new learning materials is known to be one of the most effective learning techniques [4]. Collaboration features will allow teachers to include students in the creation process.

4 Survey of Lecturer Needs

Using a usability engineering approach, we identified the lecturers' requirements for the authoring tool. We asked lecturers about their habits and practices concerning the production of lecture notes or other lecture supporting materials. Participants were encouraged to share their views on the tools used and voice their requirements and wishes for the "ultimate" authoring tool [15].

Of course, oral testimony of things users might want to do in the future are to be taken with a grain of salt [17]. Several years of co-writing lecture notes clearly showed that most lecturers are overwhelmed by the sheer number of authoring tool features to choose from without being provided with a link to the task they are trying to accomplish.

When watching the same lecturers using domain specific software (like a scientific graphing program or measurement software), one is often amazed at the efficiency with which these tools are used, however idiosyncratic and intricate they may be. Obviously software that has a strong relation to the task at hand will be easier to use. This is in sharp contrast to the "jack of all trades'' standard software solutions (e.g. office suites) which provide functionality for nearly everything, but fail to get users actually using these functions.

4.1 Materials in use

Quite often the task of providing supporting material was characterised in terms of a "necessary evil" or an unloved duty.

The materials provided ranged from none ("Visit the lecture and write down what I say, or else learn to use a library'') over various hotchpotches of assorted excerpts and papers to fully-fledged lecture notes and simple course Web sites. Another popular learning aid are copies of slides used to present the lecture.

The none option is certainly not a favourite with students. Hotchpotch versions are acceptable for students at higher levels, but trouble freshers. Everybody finds it difficult to get up to speed when required to read through dozens of papers trying to assimilate just the basic facts. e-ULE provides an easy way to sort the material and connect it with some introductory materials explaining basic concepts. Fully fledged lecture notes or even books are highly favoured by the students, but still some information needs can be better catered for by online versions.

Course Web sites were mostly used to provide downloadable versions of the lecture notes and to spread news concerning the course. A few lecturers make use of discussion groups (mostly third-party supported, e.g. MSN Communities). Discussion groups are one of the students' most wanted features, but moderating these groups can be very time consuming, thus they do not rank highly with lecturers.

4.2 Tools used

Lecture notes are almost exclusively produced using a word processor, or the TeX/LaTeX typesetting system (mainly in natural sciences, technology). One lecturer mentioned the use of a DTP system, and one was using an e-learning authoring environment. Slides were similarly done using a presentation program or the TeX/LaTeX typesetting system.

Several lecturers were using various WYSIWYG Web authoring tools to produce lecture Web sites (course information and download of materials). Nobody tried to produce entire e-learning systems - complete with lecture notes, forums and course information - using a Web authoring tool.

One university-provided tool (Sides4mi [24]) to disseminate information about a lecture on the Web received very bad marks. Mostly it was considered too much effort for too little gain. This system features a "push client" to keep students informed about their "subscribed" lectures and was rather popular among the students until the initial hype died down, and the information then became increasingly unreliable.

4.3 Most-wanted features

4.3.1 Basic needs

The ideal authoring tool should be easy to use1, stable2 and provide collaborative editing.

4.3.2 Media inclusion

The inclusion of other materials (graphics, literature, multimedia, etc.) must be easy and reliable. "Cut and paste" inclusion of various graphical material is often accomplished with scissors and glue rather than by keyboard and mouse commands. Inclusion of various media always raises the question of data formats, which overwhelms a significant number of users, so they forsake the use of media. This problem is easily overcome by providing a library for format conversion on the fly.

Mendes et al. investigated the use of hypertext authoring tools of 13 lecturers [8]. Among those, only nine made use of images and fewer than six used other media. As the authors point out, this reflects the typical contents to be found in an academic environment: text, text and text again. We do not concede this view, as it seams unlikely that only one of the projects had need of bibliographic information and only five needed links to other resources. We believe that the lack of media is mostly due to difficulties in creating and integrating these media.

4.3.3 Meta-information

Courses do not only require study materials: courses tend to come with an administrative overhead like student enrolment, times and places of the lectures/labs, etc. Some universities offer support for the distribution of this kind of information, but again, this is done via additional tools and is not integrated in the workflow. The authoring tool should provide facilities to enter information on real events.

4.3.4 Interaction and collaboration

The easy (non-technical) integration of features such as discussion forums and online chat was much welcomed and many lecturers reacted positively to the idea of involving students in the creation of materials. Some lecturers try to augment their lecture Web sites by adding third-party hosted discussion groups. Although in principle easy to use, these boards are not as tightly integrated as would be necessary for optimal benefit. It is difficult to integrate content generated on external sites, and the user is naturally dependent on the selected provider.

4.3.5 Support for link generation

Students frequently lamented the lack of hyperlinks in so-called hypermedia applications. Lecturers, on the other hand, tend to make hyperlinks only when explicitly referring to a section or resource, thus leaving the power of hyperlinks in educational materials unharnessed. The generation of hyperlinks is a rather tedious task and, as such, is best left to the computer. Interestingly enough, none of the lecturers interviewed complained about the lack of linking support: "I can make a link anytime". Only after prompting did all the lecturers concede that semi-automatic link-generation would be useful.

5 Understanding the Authoring Process

Depending on the individuals involved, the creation of academic study material can be a rather chaotic process. Complex systems - as lecture notes - involve complex, even chaotic processes that are often out of the reach of common authoring tools like word processors.

As Nkambou et al.[18] point out, curriculum development can be driven by content (content to be provided is known), course (concept of a course already exists) or material (available material is sorted and connected to teaching goals). Usually, curriculum development is an iterative succession of the three types.

Many lecturers confessed to a somewhat chaotic creation process, which often involves writing many chapters concurrently, while trying to figure out the most intuitive order. All this is done while juggling a plethora of scientific papers which might "come in handy".

Any, even the most complex, workflow can be broken down to some simple building blocks:

  1. Decide on a course
  2. Brainstorm
  3. Research
  4. Write content
  5. Define/find keywords
  6. Add additional content
  7. Organise/structure material
  8. Provide course information
  9. Organise students

The steps and the tools involved in the design of new courses are detailed in Table 2. Steps 4 to 6 can overlap (adding media, however, is frequently the last step before publishing). Steps 2 to 7 can be repeated as necessary.


Table 2. Creating a course: stages and tools used
No. Stage Tools
1 Decide on a course Unknown
2 Brainstorm Pen and paper
    Brainstorm software
3 Literature research Printed journals: filed in binders
    E-journals: stored in digital document management software or printed and filed in binders
    Web: bookmarked, saved
4 Write/reuse content Word processor
    DTP software
5 Define keywords Indices are rarely used
6 Add additional content Specialised software (e.g. CAD program, music editor, formula editor)
    Literature references: specialised software
    Interaction: Web-based discussion group server software
7 Organise/structure material Word processor?
8 Provide course information Lecture Web site: Web design program
    FTP program
    University provided tool for lecture Web sites
9 Organise students Email program
    University provided tool for student data manipulation


6 e-ULE -System Overview

A university level e-learning system should reflect the academic nature of the contents to be taught as well as the limited budget most universities are confronted with. It also must take into account that most university teachers are neither trained information scientists nor educationalists. Thus the system must provide an easy-to-use authoring system and a process to guide the lecturer towards the creation of a usable educational hypertext system.

The aim of the e-ULE system is threefold:

  • Creation of an educational hypertext design process (EHDP) and an educational hypertext markup language (e-ULE_doc).
  • Generation of a server backend to handle adaptive and collaborative features as well the generation of personalised/specialised versions of the material.
  • Generation of an authoring tool geared towards the creation of university level hypertext.

6.1 Comparison with Intelligent Tutoring Systems

e -ULE differs from conventional Intelligent Tutoring Systems (ITSs). These systems are time-consuming to produce and often require an entire team of system experts [3,12]. ITSs are still too much of a research playground to help individual lecturers, especially those from a non-technical background. Existing systems mostly focus on relaying of "procedural" information, like programming languages or mathematics.

An e-ULE project can be handled by a single lecturer, or small group. The time required to produce an e-ULE project does not differ vastly from the production of conventional lecture notes.

While ITSs are able to adapt completely to the students' learning situation, such behaviour might not be sought in universities, where students are expected to acquire knowledge autonomously.

Adaptive in an e-ULE sense means that the system will recognise the student and provide a custom interface to the contents of a project. The system will create a personal history, markup the most frequently used pages, and notify the student about changes to pages recently visited. The student can create custom views of the material available and even have the system create paths to selective chapters of the material, which will include all necessary prerequisites. e-ULE's open architecture also allows students to participate in the creation process, thus deepening their knowledge of the domain.

Most ITSs assume the lecturer possesses all knowledge and a master plan of the connections and builds the ITS accordingly. In the real world lecturers have to look up material (quite a lot of time), or completely rearrange the path of their lecture, or may want to use legacy materials. The inclusion of legacy documents is no problem in e-ULE. The lecturer is also supported in gathering information and in structuring the project.

6.2 System components

e -ULE consists of three major building blocks: server, authoring tool and a student client. These building blocks are used to create, distribute and view e-ULE documents. The various materials (e-ULE documents, papers, meta-information) pertaining to a course are collected in an e-ULE project. Figure 1 depicts the various parts of the e-ULE system.

figure 1

Figure 1. Components of the e-ULE system plus their implementing technology. Most e-ULE features are located on the server, with a slim authoring client and a standard Web browser on the student side

6.2.1 Server

For the sake of usability most of the functionality is integrated into the server. Once installed, an e-ULE server can serve a whole department, or even a university.

The server's tasks include:

  • serving the content
  • providing interactivity (forums, chat, annotations)
  • providing adaptive features
  • enabling collaborative editing
  • ensuring version control
  • student and course management

The server is built on an Apache - Jakarta - Cocoon platform.

6.2.2 Student client

As most of the functionality is handled by the server, a common, up-to-date Web browser and PDF viewer are sufficient. This ensures that a wide range of common data formats can be easily integrated in e-ULE projects. Expert data formats (like CAD formats) might require additional plug-ins.

The student can choose among different representations, including on-the-fly generation of "sub-notes"3, as well as offline and print versions. These options ensure the appropriate material can be used for learning, exam preparation, reference or just to gain an overview of what the course is about.

6.2.3 Authoring tool

The multi-platform authoring tool is realised as a "slim client". Data are not stored locally but on the server, which also handles version control and collaborative editing. At the moment it requires a permanent connection to the server, which should not be a problem in the intended environment, but an offline mode might be integrated in further releases. Section 7 explains the features of the authoring tool in detail.

6.2.4 e-ULE documents

e -ULE relies on XML as the document format. XML is ideally suited to the type of semi-structured content usually found in lecture notes. Due to the common XML format it will be easy to convert content to and from other XML based e-learning formats like SCORM [22], IMS [6] or EML (Educational Markup Language) [19]. Additionally a host of other XML-based languages like MathML (Mathematical Markup Language) or CML (Chemical Markup Language) can be used.

6.3 Core concepts

e -ULE shall provide usable and engaging support for all types of courses. This requires an equally usable authoring tool, that frees the lecturer from the burden of technical details so he can focus on the actual content.

6.3.1 topics and keywords

e -ULE centres around topics. topics are information "atoms" which cannot be subdivided4 meaningfully (compare indexing [3] and atomic concepts [26]). A strict process ensures that all meta information necessary to build structures automatically is available. Every topic comes with:

  • a title, an optional subtitle, keywords, an abstract
  • prequel and sequel information
  • level information
  • optional extends/requires information
  • optional additional content

topics are contained in blocks and are linked by prequel and sequel identifiers, thus resembling book-like structures (cf. [12]). In addition, topics come in three flavours: essential, important and additional. This level information allows the student to access to the core elements of a lecture faster. Relations between the levels of the topics are marked by extends and requires identifiers.

Abstracts of topics are used to generate a padded table of contents (the often wished for "90-minutes notes" or "executive summary"), which are suitable for a first overview of the lecture.

Keywords denote the actual topic and semantic variations. This information is used by the link-builder automatically to create a glossary, an index and appropriate hyperlinks.

6.3.2 Semi-automatic generation of hyperlinks

Hyperlinks set hypertext apart from ordinary textual material [11,13]. Unfortunately the manual generation of hyperlinks is quite tedious, as can be seen by the number of "brochureware" or "electronic page turning" [7] Web sites. Educational hypertext is no exception.

The main feature of e-ULE is the automatic generation of hyperlinks. A process similar to a spell checker is used to build custom dictionaries of lecture keywords. These keywords are used automatically to generate hyperlinks and custom views of the material.

6.3.3 Collaborative editing

As many courses are presented by a group of lecturers, collaborative editing is another crucial feature. In line with the constructivist learning theory, lecture notes could even be written in cooperation with the students. Teaching is commonly considered (see e.g. [4,20]) one of the most effective ways to further deep understanding of a subject matter. Using the e-ULE students can participate in the creation of the lecture notes and thus deepen their knowledge. Student participation can be realised by annotating existing material, or by completely integrating them as additional authors in the collaborative authoring environment.

7 e-ULE Authoring Tool

Any learning environment that aims to create usable, interactive materials for students must provide an equally usable tool for authoring these materials. As pointed out by Shackelford [23]: "typically, such software does little or nothing to support teachers[...]". e-ULE supports the lecturer in all parts of the workflow (see section 5).

figure 2

Figure 2. e-ULE's support for the different stages in the authoring process. Steps 2 to 7 mark the core process, which will be repeated. Not all steps need to be accomplished in each iteration

Looking at various existing authoring systems, one can find many different approaches: WEAR [12] and AHA [1] use HTML files which are augmented with meta information on the server side, while Interbook [2] relies on specially structured MS-Word files plus concept-based annotations. WebCT [5] provides a server based HTML forms/javascript interface.

After thorough evaluation we decided on a slim authoring client to improve speed and reduce administrative overhead. Currently we are evaluating a Java client and an interface based on the Mozilla toolkit.

7.1 Project start

A new project comes with one automatically generated author/editor (the person who starts the e-ULE project). The main editor can appoint any number of additional editors (with full control over the project) and authors (rights can be assigned).

Projects can be built from scratch or can be derived from other e-ULE projects. At the start of a project server space is allocates  and several new databases are created for the collection of ideas and documents. Existing idea and document repositories can be attached to the new project.

e-ULE projects can be password protected or open to the public. Guest accounts assure that copyrighted material is hidden from unauthorised view (e.g. papers via site license).

7.2 Brainstorm mode

Modern word processors frequently provide an outline mode, to facilitate the generation of structured, tree-like documents, thus forcing the author to think in chapters, sections, subsections, and so on. e-ULE goes one step further and allows for a "semi-chaotic" authoring style by providing a brainstorm mode and an idea repository. At any given time the author can jot down a new idea or start a brainstorm session (where new ideas are automatically entered into the system at the press of the enter key). Ideas are kept in the idea repository. Ideas similar to the topic at hand will be presented to the author for easy inclusion in the current topic. The contents of the idea repository can be browsed and searched at leisure.

7.3 Literature research

Writing lecture notes usually involves a lot of research on the topic in question. Support for the research phase is a crucial part of the authoring tool.

Even though the Web is now a major source of information, there are hardly any tools available to harness the power of the Web. Many scientific journals can be read online, and many more are to follow, but handling of this wealth of information proves rather tricky. Every publisher has its own system - incompatible with others. Once the articles are safely downloaded, the real hassle begins. Articles cannot efficiently be organised, categorised and processed. In the end many users capitulate - articles are printed and filed in binders.

e-ULE will provide extensible support for literature search and management. Articles can be searched, downloaded, indexed, annotated, categorised and quoted without ever leaving the authoring tool. The same applies to any type of Web-based information. Web sites can be bookmarked and stored offline for further reference. Even if the site vanishes from the Web, the information will still be accessible.

Information which was digitally received while working inside the e-ULE environment will be stored in the document repository on the server. An indexer ensures that articles containing the same keyword as a topic under way will be available to the lecturer.

Legacy documents (word processor file, PDF, Postscript, etc.) available in digital form can be added to the document repository at any time. If the format cannot be processed by the indexer, the lecturer can specify a description and optional keywords.

7.4 Adding content: text, inline objects

Text can be added as with any modern word processor. e-ULE supports all commonly used content types like lists, figures or tables. We intend to provide support for formulas and chemical symbols as well, as soon as modern browsers are able to interpret MathML and CML correctly. At the moment support for formulas is limited to graphics or LaTeX (converted on the server).

Content also includes inline objects, most commonly graphics, sound, animations and video but also expert data formats like CAD files, destined to flow with the text. Data types common to the Web are automatically detected and integrated; special types require either user intervention (specifying a link for the appropriate browser plug-in) or are handled as references on external objects.

Visual markup is discouraged; in fact it is limited to italic font to emphasise certain parts. Allowing free markup would interfere with visual cues for keywords (etc. ) provided by the system and furthermore distract the author from the contents. A range of logical Markup like quote, keyboard input, heading is available, however.

7.5 Defining keywords: common sense checker

The common sense checker lays the foundation for the automated link-builder. In a process similar to spell-checking, the authoring tool will mark all unknown words. Actions on unknown words include: correct spelling mistake (choose from a list of recommendations), mark as new keyword, mark as a deviation from a known keyword (choose from list of similar keywords).

New keywords come in two flavours: prerequisites, which are used to generate a special list of "should know" items and real keywords. The real keywords are immediately associated with a new topic and a glossary entry. Every topic can be manually associated with several keywords. These entries can be selected from each occurrence inside the text.

The common sense checker, of course, relies on user input to point out useful keywords. This cannot be avoided, as the very same word could be a keyword or just an ordinary word, depending on the audience. An integral in a freshman's course on mathematics for economists has a different meaning from an integral in postgraduate course of particle physics, for example. In addition to the words detected automatically, the user can convert a word into a keyword at any time.

7.6 Additional content

e -ULE features typed links (compare [26], teaching operations [3]) to include additional content like examples, exercises and solutions, assignments, bibliographic information, hyperlinks and external material. Server modules add interactivity to the lecture notes. Currently available modules are discussion groups (forums, chats), annotations and online assignments (with peer review option).

7.6.1 Discussion groups

Communication is a vital aspect of learning. Since the aim of the e-ULE system is to deepen student-lecturer communication by reducing the pure information transfer aspect of the lectures, there is room for communication in the lecture hall. Still, there will be the need to communicate electronically: in the forums students can talk to each other, to the lecturer or to tutors. Different communication needs are served by different types of communication channels: asynchronous, newsgroup-like systems (forums) and synchronous chat systems.

Server-provided discussion groups and online chats can easily be integrated and may belong to one single topic or to a group.

7.6.2 Annotations

Each topic can also be annotated by students. These annotations can be either public or private. Annotations are another means of communication. In a noticeboard manner annotations can serve as platform for collecting additional information about each topic, like errata, implementation examples, notes about practical applications or additional Web links.

7.6.3 Assignments

It is also possible to add assignments and have the results submitted electronically. Submissions could then be distributed for peer review: students would review the work of another student. The resulting reviewed assignments can be attached to the e-ULE projects.

7.7 Organisation and structure

As a lecture's content tends to be sequential (usage of the lecture notes is not necessarily sequential), e-ULE provides a central TOT (Table Of topics) to group and (re)arrange topics. topics can be "stacked", providing information in three levels of difficulty (essential, important, additional). Flags mark topics as draft, ready for review, reviewed or completed (ready to be published). Whole topics, or tasks on topics, can be assigned to a specific co-author.

The all important (and universally disliked) to-do lists will be taken care of by e-ULE: if a topic is missing compulsory parts (like the abstract or a level mark), the need for completion will be flagged. Tasks can be assigned to members of the authoring team. e-ULE will also help in planning a new course. The lecturer can formulate goals and assign topics to the team - complete with a timeframe for each topic/goal - and thus build a course table from it.

7.8 Course information and administration

The information section includes information about the lecture and the e-ULE site. This may include: lecture times/dates and places, an introduction to the e-ULE client, contact information, technical support, and other relevant information.

An e-ULE project can be public or restricted to a certain group of students. Usernames and passwords can be automatically created, or users can be authenticated against other servers.

7.9 Visual formatting

Being Web-based, WYSIWYG is not a key concept in e-ULE. The author can choose from various styles, which can be adapted by a few simple choices. On-screen display varies naturally with the platform and the choice of browser. Printable  PDF output is rendered via LaTeX. As the system is open-sourced and transformation is achieved by XSLT, it is possible to create custom styles.

8 Summary and future plans

e -ULE helps the lecturer in all phases of course development by providing an integrated environment for: literature research, authoring, administration, and communication plus a strong guiding process.

On the student side e-ULE features support for different learning phases and learning styles including the possibility of tailoring the materials to individual preferences (specific views, annotations, adaptive features).

The system will be open-source and released under the GNU General Public License, building on open standards and the wealth of open source software already available.

Currently e-ULE is under active development and a number of lecturers have volunteered for usability testing of the prototype.

References

1 De Bra, P. and Calvi, L. (1998) "AHA: a generic adaptive hypermedia system". In Proc. 2nd Workshop on Adaptive Hypertext and Hypermedia

2 Brusilovsky, P. (1998) "Adaptive educational systems on the world-wide-web: A review of available technologies". In Proceedings of Workshop WWW-Based Tutoring

3 Brusilovsky, P. (2000) "Course sequencing for static courses? Applying its techniques in large-scale web-based education".
In Proceedings of the 5th International Conference on Intelligent Tutoring Systems, LNCS No. 1839 (Springer)

4 Chi, M., N. de Leeuw, et al. (1994) "Eliciting self-explanations improves understanding". Cognitive Science, International Journal of Artificial Intelligence in Education, 18(8)

5 Goldberg, M. W., S. Salari, and P. Swoboda (1996) "World wide web - course tool: An environment for building www-based courses". Computer Networks and ISDN Systems

6 IMS (2002) IMS specifications
http://www.imsproject.org/

7 Jones, V. and Jo, H. (1998) "Interactive multimedia based on learning theories to enhance tertiary education".
In ICCMA'98 Conference Proceedings

8 Mendes M. and Hall, W. (1999) "Hyper-authoring for education: a qualitative evaluation". Computers & Education, 32:51-64

9 Mitchell, C., T. Dipetta, and J. Kerr (2001) "The frontier of web-based instruction". Education and Information Technologies, 6(2), June

10 Montelpare, W. J. and Williams, A..(2000) "Web-based learning: Challenges in using the internet in the undergraduate curriculum". Education and Information Technologies, 5(2), 85-101

11 Moulthrop, S. (1991) "Beyond the electronic book: a critique of hypertext rhetoric". In Proceedings of the 3rd Annual ACM conference on Hypertext

12 Moundridou, M. and Virvou, M. (1992) "Authoring and delivering adaptive web-based textbooks using wear". In IEEE International Conference on Advanced Learning Technologies: Issues, Achievements, and Challenges - ICALT

13 Mynatt, B., L. Leventhal, et al. (1992) "Hypertext or book: Which is better for answering questions?" In Conference Proceedings on Human Factors in Computing Systems

14 Naber, L. (2001) "Hypertext support for learning phases". Technical report ifs01-36, Vienna University of Technology

15 Naber, L. (2002) "Authoring tool requirements analysis". Technical report ifs02-33, Vienna University of Technology

16 Naber, L. (2002) "Student motivation survey". Technical report ifs02-31, Vienna University of Technology

17 Nielsen, J. (2001) "First rule of usability? Don't listen to users". Alertbox, August 5th
http://www.useit.com/alertbox/20010805.html

18 Nkambou, R., C. Frasson, and G. Gauthier (1998) "A new approach to its-curriculum and course authoring: the authoring environment". Computers & Education

19 OUNL (2002) EML specifications
http://eml.ou.nl/

20 Renkl, A. (1997) "Learning from worked-out examples: A study on individual differences". Cognitive Science

21 Schell, G. (2001) "Student perceptions of web-based course quality and benefit". Education and Information Technologies

22 SCORM (2002) Advanced distributed learning initiative.
http://www.adlnet.org/

23 Shackelford, R. (1990) "Educational computing: Myths versus methods. why computers haven't helped and what we can do about it". In ACM SIGCAS Computers and Society, Proceedings of the Conference on Computers and the Quality of Life

24 unikat. Sides4mi - das personalisierte informations- und kommunikationssystem für die lehre.
http://www.lzk.ac.at/sides-4mi.html

25 Watson, D. (2001) "Pedagogy before technology: Re-thinking the relationship between ict and teaching". Education and Information Technologies, 6(4)

26 Wu, H., G. Houben, and P. De Bra (1999) "Authoring support for adaptive hypermedia applications". In Proceedings of ED-MEDIA '99

Notes

... use1
devoid of function with no relevance to the task at hand
...stable2
Everybody(!) working with a word processor ran into trouble at least once when editing lengthy documents: documents were either mutilated or even completely lost and had to be reconstructed from earlier backups
... "sub-notes"3
A sub-note is a part of a lecture note/e-ULE project that highlights a particluar aspect of the material available and includes all necessary prerequisites
... subdivided4
Can be thought of as subsections, or sub-chapters