Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The number of multimedia products with text, graphics, sounds, animation, and movies is increasing. Many of them are CD-ROM titles and there were more than 2,000 of these in 1991. Many of them have some kind of linking to navigate through.
3.1 Electronic Books
How do you read printed books? If you read novels, you will start with the first page and read through sequentially to the end. In this case, particular parts of the book do not need to be linked together. But what if you use an encyclopaedia? How do you read an encyclopaedia? You will not read it from the first page. You will first look at the table of contents and then turn to the page with the item you are interested in. Or you will look at the index at the end of the book and go to the pages to which you are referred. In an encyclopaedia, there are simple linkages of blocks of pages by contents, index, and words. The page where a specific item is explained might include references to other parts of the book or to other books. The reference to other books can be thought of as links to external sources of data. Anyway, Hypertext that can include sophisticated linkings might not be very useful for software products where we can so easily anticipate how users will use the data they contain.
Now, let us examine two electronic books where we can see a simple linkage of data and multimedia such as audio and pictures. One is a CD Book for children and another is an electronic encyclopaedia. Figure 9 shows pages from Peter Rabbit (DISCIS Inc.). The left page shows text and the right page shows a picture (on the screen it is in colour). We can flip pages forward and backward using the clips at the bottom corners of the pages. Clicking any part (object) of the illustration, we can get the word that explains the object and can hear it pronounced. We can hear the pronunciation of any words on the text page and can also hear their definitions. This means that objects on the picture pages and words on text pages have links to the text and the voice. This CD Book is an electronic edition of the famous book and we need not navigate through it, but there are simple links between the objects on the pages and the data hidden within the pages.
Figure 10 shows the screen of the Grolier Encyclopedia, which is an electronic edition of the printed encyclopaedia. There are more than 30,000 articles in the book and we can browse through them by title index in a search menu at the top of the screen. But if we need to read articles about some specific word, we must have some navigation tool because there are so many items and words included in this book. The encyclopaedia has a search system based on words. Figure 10 also shows an example of a word search, where we entered the phrase "UNITED NATIONS" and 260 articles were found.
This encyclopaedia has more than 3,000 colourful pictures and many sounds including music and bird's singing that are linked to articles. We can enjoy beautiful paintings, scenery, animal pictures, etc., and we can also enjoy listening to a symphony orchestra and bird's singing. Figure 11 shows the article about Beethoven that was retrieved by word search. The screen shows the text article, a picture, and audio. We can see menu icons in the upper corner of the article through which we can see pictures or listen to sounds. We can use arrows to go to the next occurrence of the word we want to see. We click the camera icon to see the picture that is linked to this article. We click the headset icon to listen to the audio that is also linked to this article.
These two electronic books do not have sophisticated links, but we can enjoy multimedia (pictures and sounds) and we can easily access the articles we want to read.
3.2 An Electronic Magazine
Authorware Inc. (now MacroMedia Inc.) tried to publish an electronic edition of MacWorld magazine published by the IDG group. Figure 12 is the first title screen of that magazine. Though it was a trial that included only part of the printed copy, it shows us the future electronic magazine. Figure 13 shows one page from that magazine. At the left there is paging and we can go to a specific section or flip from one page to another. The text inside the screen has highlighted words in it and those highlighted words have links to notes and pictures. Figure 13 shows the picture revealed when we click the highlighted word, "field of view." The pictures linked to the word are animated, and if we move the camera on the picture by Mouse, then the camera view will change to a different angle. It is also linked to another article. If we click the arrow, we can, for example, go to another article that explains "ART BEAT."
This electronic magazine has text, pictures, animation, and sounds. We can put bookmarks and memo on any page of the magazine. There are also monthly magazines based on CD-ROM, like Nautilus. Many of them have some sort of linking and use multimedia.
3. How useful is hypermedia for business people?
Hypermedia might be interesting from the technical point of view, but it must also have a practical application to be used in the business environment. Why do we need links of data blocks and why do we need Multimedia? Because there are so much data that we need some tool to get at the relevant information; because information limited to texts and figures is not fully satisfying to management. If they can see information in graphical form, in animation, in movies, and if they can listen to spoken information, then it might give them more satisfaction.
In order to process the flood of data, there exists a variety of database softwares. Data are stored in databases in a structured way and we can retrieve meaningful information from databases based on key subject fields. But, traditional database software needs a rigid file structure on which to store and organize the data to be retrieved. If we could place all data on a specific framework, then it would not be a big problem to have access to relevant information from the database. But the difficulties are that there is such a variety of data in the world that we cannot structure all of them in a specific framework. We need new tools in order to get meaningful information from unstructured piles of data and the linking function of Hypermedia will help us navigate through unstructured data.
There is another problem. If we have an information intermediary at hand and can ask this person to get information for us from the database, then it is rather easy for us to get information. But in order to let the intermediary know our information needs, we must discuss them, which might take rather a long time if that information is unstructured. Even so, if we do have easy access to professional help, then we are very happy; but in fact there are not many people who are able to enjoy this help.
3.1 Structuring Data by Outline Functions
When we write a long report of 50 pages, it is difficult to write from the beginning to the end. We need some framework or structure to provide an overview of the full document. Outlining is a good tool for giving structure. We can write a long text based on the structure and can then easily restructure the document. The outline function was first introduced in an IBM-PC software called Thinktank (Living Videotext Inc., now a division of Symantec Corporation).
Figure 14 shows the structure of a document with four levels of topics (the software used is Acta of Synmetry Corporation). The top level is identified by topics headed 1.0 and the next level is identified as 1.1, 1.2. The next follows as 1.1.1, 1.1.2, 1.2.1, 1.2.2, etc. We can hide lower levels of topics with the text. In figure 14, a black arrow next to an item head shows that there are hidden items under it.
For example, topic 1.1.1 has a black arrow and does have lower level topics hidden under the screen. If we can show only the topics we want to see at the moment and can hide other parts or levels of a document, it is much easier to see a specific part. Of course, if we move a topic to another place, then all the subtopics move with it to the new location. We can also change the level of topics. For example, if we change topic 1.1.1 to a higher level, then the topic numbers will be reshuffled to reflect the new structure.
Today, many word-processing softwares have this kind of outline function. Other softwares also have outline functions. Figure 15 shows a presentation software (Persuasion by Aldus Corporation). In this figure, several topics are listed under an outline structure. If we change the screen from outline view to slide view (figure 16), the topics on the slide have a structure of topics in indentation.
It is interesting to point out that Excel (a spreadsheet software of Microsoft for Macintosh and Windows) also has the outline function. Figure 17 shows sales data on a worksheet, and there are so many items on it that we cannot see them all on one screen. In order to see another part of the worksheet, we need to scroll vertically or horizontally. But as you see in the formula menu, there is an outline function. If we use it and give a specific structure to the data based on the formula in the cells, we will get the outlined view of the worksheet. Figure 18 shows the screen after we give it a structure using the outline function. As you see on the left side of the figure, there are five levels of structure, and one can set the level of data that should be shown on the screen. If you compare these two figures, you can understand the usefulness of outlining for navigating the whole worksheet, especially when it is very large.
Two more softwares also have the outline function. Figure 19 shows a screen of Inspiration (** Inc.). Using this software, we put ideas on the screen based on various symbols. We can easily combine ideas into more meaningful ideas. In the figure, there are ideas for formulating a strategy to increase profit. There are ideas, sub-ideas, and sub-sub-ideas. We can see a hierarchy of ideas on the screen and a tool to symbolize ideas. If we change the style to the outline mode, then we can see the outline form of the ideas (figure 20). If there are many ideas and they have multi-level relations, it is often helpful to have them in outline form.
Project management softwares help us monitor the time, cost, and human resources for accomplishing a big project that includes many tasks. One task may have several related sub-tasks and the sub-tasks may again have sub-sub-tasks. So there is a hierarchy of tasks in a big project. When we monitor a project, we cannot see and do not need to see all the tasks at one time. So if we can give it structure and can focus on one specific part at a time, it will be helpful. Figure 21 shows a screen of a project management software with a structure of many tasks (On Target by Symantec Corporation). The + symbols at the left of the task names show that there are several subtasks hidden under the screen. As shown at the top of the screen, there are several commands in the Outline menu for managing the structure of the tasks.
The outline function does not have the linking function, but it is very useful for business people when structuring documents, ideas, or data on a worksheet.
In a spreadsheet, we can link a part of a sheet to another part of the sheet or link a graph to specific data. It is a macro assigned to an object on a worksheet. Figure 22 shows an Excel worksheet where macros are assigned to buttons that show graphs. If we click the Japan button, then a macro assigned to it shows the graph based on the specific data. Macros and buttons perform the linking functions on a worksheet.
3.2 The Use of Personal Information Manager
There are many data that cannot be placed on the structured field of a regular database. For example, topics in a newspaper may be much longer than the width of a field set in a database. This kind of data is called free format or unstructured data, and there are several softwares that handle this well. PIM (Personal Information Manager), Agenda (Lotus Development Inc.), and GrandView (Symantec Inc.) are well-known softwares in this field. Figure 23 shows a screen of Agenda, where items with the keywords are automatically linked to the words shown at the right in each category with the heading of "software," "company," etc. After getting many data that are linked to words, we can retrieve items linked with the words for which we want information.
Another useful software for free format data is Thought Pattern (*$ Inc.). We can easily give a tab(index) for a text and use tabs to get information from a free format database. Figure 24 shows a screen of Thought Pattern, where we select a word as a tab for this text data. The Tabs menu shown at the top includes several commands to create tabs, and if we use Cross Index, Thought Pattern automatically compares registered tabs with the words in a text data and makes links to the words appearing in the text data.
We can easily set tabs for retrieving data by selecting tab words in the tab lists, as shown in figure 25. There, tabs are grouped in a hierarchy, and we can use outline mode to show the level of tab we need to use for retrieval.
4. Executive information systems
There are softwares called Executive Information Systems (EIS). By using these, managers and their staff who are not computer professionals can get meaningful information for themselves from piles of data stored on different databases. Data can be arranged in a hierarchy from general to detailed levels. Managers can start searching information at the top of the hierarchy and go down to lower levels to make clear what information they want. This process is called "drill down."
Data may be stored in different databases. For example, sales data may be stored on a dBASE file, accounting data may be stored on a Lotus 1-2-3 file, and customer data may be stored on an Oracle file. Some files might be stored on a PC, and others might be stored on workstations or mainframes.
The EIS softwares enable us to navigate this myriad of information with minimum help from intermediaries. Figure 26 shows a screen of EIS software (Lightship by Pilot Executive Inc.). It shows icons, and there is a hand pointer on the second upper level icon. A hand on an icon shows the presence of linked information at a lower level. By clicking this icon we reach the lower level. Figure 27 shows the result where there is a map of the United States, which again has several icons on it. Moving the Mouse pointer to one that has linked information again changes the pointer into a hand pointer. In this figure, the Mouse pointer is on the "Profit Margin by Region" icon and we can drill down again one additional level. Figure 28 shows a graph of "profit margin" at that level.
The problem with this kind of software is the organization of data to provide meaningful links. Of course, there should be development tools built into it. Figure 29 shows a screen for the development stage. We can see several menus for this stage.
For example, the tools menu includes commands like charting, drawing, text writing, image processing, etc. We can easily make a specific screen that shows data from the database, texts, graphs, images, sounds, etc., and we can link that screen to other screens at any level of the structure.
One difficulty of EIS is how to define information needs. Information comes from a database by filtering data according to specific criteria. One of the most popular tools for defining an information need is a special language called SQL (Structured Query Language). There is an EIS software that includes many interesting functions. Figure 30 shows a screen of the EIS software (Forest & Trees, Channel Computing Inc.). There are several windows that show specific information displayed on it such as a table, a chart, or a single number. These windows are called "views," and we can define more than 2,000 views in any way we choose. Some views show information based on a specific database and some views the result of processing the numbers contained on several views.
We can use SQL to define our information needs. For example, figure 31 shows the definition of the "Best Sellers" view shown on figure 30 and displays a list of different database softwares. When we define a view, we first select the database from which we want to get information. The Forest & Trees shows the list of tables (files) that are created by the specific database software, and we can select a specific table. Figure 32 shows where we can define the information need by using SQL. This illustration shows an SQL program to get the best-selling products based on "fish" and "sales" tables. SQL can be used for any files listed in figure 32. There are R:BASE, dBASE, Lotus, Excel, Btrieve, etc., and we can add drivers for mainframe databases like DB2, FOCUS, etc. The same SQL syntax can be used for any kind of file, and it is rather easy to write programs with the query assist shown on the lower part of figure 32.
In Forest & Trees, there is a drill-down function by which we can navigate the information hierarchy. Figure 33 shows the tree structure of a view named "current ratio." This view is obtained from the two views called "current assets" and "current liabil" (liabilities). Those two views then are created from views at a still lower level. Finally, the lowest level views in the tree get information from specific databases. We can also see many icons at the top of the screen that may be useful to users. from which we retrieve data may be updated at different time intervals. For example, a sales file may be updated every day, an accounting file every week, and a customer file every month. Then, in order to get fresh data, we need to keep in mind the date and time of the updating schedule for all the different files, and at the scheduled time we must access the updated file to get a new result. But this procedure is tedious work. Forest & Trees helps us by automating this work. Figure 34 shows the schedule-setting window for "current ratio" view. In it, we set the schedule for every Monday at 12:00 noon. Then Forest & Trees will automatically access the data files as scheduled.
One of the nice things in this software is its scheduling function. Data files
Why do we need information? Because we need to analyse it to see whether something is happening or not. If something troubling is happening, we do something to take care of that problem. For example, if we see a "current ratio" view and find it is lower than 2, we must prepare against being short of cash. But this situation may occur only once in 10 times. Almost all the time, we simply access the database to verify that the current ratio is higher than 2 and feel happy. If Forest & Trees can monitor the database and let us know when the current ratio is lower than 2, it is a great help to us. EIS should perform that kind of service.
As explained before, each view is linked to the database through SQL programming and linked to other views. Each view can be assigned a specific condition, and if the view is in that condition, it automatically signals the need for corrective action. Figure 35 shows the "set condition" screen, where we set condition in which, if the current ratio drops under 5, then Forest & Trees should pop up an alarm message on the screen to signal that the critical condition is present. We can arrange for the view to perform many more tasks.
Hypermedia enables us to navigate through the sea of information that is composed of text, numbers, graphics, pictures, sounds, and movies. It is interesting to experience movies, music, books, encyclopaedias, etc., interactively. The key point of Hypermedia is linking information, and it is very important for business people at the office to have access to useful information from databases. 1 described several Hypertext and Multimedia softwares available at PC shops. I also described PIM and EIS softwares, which have the linking feature and are helpful to business people. We also saw that outline functions are useful to help structure a document, a worksheet, an idea screen or a task list. The outline function does not have an explicit linking capability, but it does enable us to structure data and helps us to navigate the data.
Several issues were raised by H. Yamada during the discussion. One, dealing with methodological and related socio-technical problems, was the importance of the analytical procedure vs. the prevalence of the synthetic approach in understanding man-machine systems.
G. Johannsen explained that the synthetic approach is indeed rather ambitious, but that many components have already been adequately analysed, while others have not. The emphasis was on setting a framework that includes more influencing factors than are usually considered in working with man-machine systems and especially more human agents, who need to understand the process. This means that a holistic framework should be adopted in order to recognize the kind of strategies and tools that need to be developed for better understanding and use. He said that this is done with respect to improved training for acquainting people with technical man-machine systems, dealing with knowledge-based information access by means of simulation and self-learning tools. He emphasized cultural aspects of such tools and the need for future research and socio-technical development.
Another aspect discussed was the "hostility against technology." H. Yamada said that this may also be related to the tendency of the scientist to analyse too much without integration, making science obscure and contributing to creating mistrust towards science in the general public. M. Dierkes advanced the view that at present there is less of this mistrust: the public increasingly understands that there are costs and benefits bound to the use of new scientific results and of the science-based technologies. He said that the "visions" of future technologies, as presented here, are an important social and international component in deciding on the "desirable" directions for technological development. Public discourse on such visions would help to make better selections and to avoid some of the negative side-effects that have occurred in the past. In this relation, 1. Wesley-Tanaskovic stated that the adoption of the broad scope of man-machine systems, as presented here, and of the concept of "socio-technical development" was seen as a possible way to bridge the existing "gaps," especially the knowledge gap between various social groups and between nations as well.
Commenting on "human-centred design," W. Rouse said that among the various "stakeholders" interested in transfer of information, the most difficult to deal with is the aspect of the "producers," e.g. researchers and engineers. Incentives, and related structural changes, are needed to assure that they make usable and useful contributions to the world's knowledge base, without, however, constraining their creativity. New information technologies, as seen here, are means to enhance information stores, not just to fine-tune the information systems. He further stated that there are many examples of successful applications of the new information tools and telecommunications for rapid transfer of requested scientific information between the continents. This makes the scenario of the 1990s quite different from that of 20 years ago, when the first world science information systems were designed.
In relation to this aspect, M. Almada de Ascencio reiterated that scientific information systems were first thought of as being for the researchers themselves and were set up by them. Not of primary concern were other sectors, such as local government, business, and industry, which all need information for problem solving and, often, scientific and technical information as well. The proposed "human centred design," by providing more adequate information support, would seem to her to be especially useful for the developing countries building their own systems.
Speaking of hypertext systems, their potential as educational tools was questioned by J. Alty. Responding to him, N. Streitz pointed out that not every hypertext system is at the same time an educational learning environment. In order to provide the educational experience, one has to combine the hypertext/hypermedia environment with didactic strategies. In addition, he said, one has to take into account who the learner is, the subject domain, etc.
At the end, in relation to the comprehensive review of currently available Personal Hypermedia Systems, demonstrations were made of some new industrial developments by Fujitsu Ltd. of Tokyo, Japan, which could serve as examples.
