Some Issues of Transition and Information Interpretation in the World of Digital Libraries

1. Framework

Traditional Library	Digital Library
Come-in (physical access) Document-based (monomedia) Just-in-case (depository) Temple of knowledge Data and information Information system Satisfy-the-patron Source and destination Mono and policultural Facilitator of change	Plug-in (virtual access) Content-based (multimedia) Just-in-time (upon request) Exploratorium Data, information, knowledge Expert system Delight-the-client Point of departure Truly multicultural Catalyst of change

2. Information Interpretation in the Digital World

1. Users of in formation easily "anchor" around some known (mostly numerical) facts, regardless of their relevance; For example, a professor, before examining a student, checks his record and finds out that most of his grades are "B". If this fact does influence professor's grading, regardless of the knowledge shown by the student, than we say that the "anchoring" takes place. The information one retrieves, even though it may be totally irrelevant, has affected the decision. The anchoring happens more easily to those individuals who are either not sure, or don't have clear picture of a problem, or don't understand its nature. The unstructured and ill defined "digital" world of Internet is an ideal environment for anchoring to take place. 2. Users are often insensitive to prior probabilities of the expected events; For example, a passionate smoker reads a newspaper article and finds out that clinical tests prove six times higher probability of lung cancer in smokers. However, he does not decide to quit smoking. Most individuals 'in principle' appreciate the law of large numbers. But their behavior is more often ruled by the "it-works-for others-but-not-for-me" principle. Therefore, in order to satisfy the user's needs more adequately, an objective interpretation of facts and statistical data should accompany the actual data. It is a difficult task to accomplish in the framework of a spontaneously generated digital information world.

2. Users fail to appreciate the sample size. Therefore, they attribute the same probability of occurrence to events, even though information about these events may come from samples of very different size; For example, when a person decides to buy a new car, his decision is usually more influenced by opinions of a couple of close friends (small sample), than by official tests or a poll of opinions, published in a major professional journal. This attitude somewhat resembles the insensitivity to prior probabilities. Users who fail to appreciate the sample size, often make decisions with confidence. That is, in case of small samples information sources are known, while otherwise, such sources tend to be anonymous. In the digital library environment this becomes an issue of ever growing importance, since the "sample" of offered information depend s on availability in digital form, and not necessarily on relevance.

3. Most people prefer redundant information sources to independent information sources; For example, a sovereign surrounds himself with flattery of his courtiers. Consequently, he is not going to favor criticism by any of his subjects, even though it may originate from their best intentions. Redundancy among information sources decreases trustworthiness of the obtained information. Suppose, for example, that a manager chooses to read only the reports of the departments in which increase in sales took place.These reports will not teach the user of such information much about all the departments. Independent information sources provide the whole array of viewpoints. They often replace the safe pleasure of one-mindedness by an uncomfortable collection of doubts. Bertrand Russell, the British philosopher supposedly said: "Those who know a little, radiate with self-confidence and assurance. Those who know a lot, are troubled by dubiety." Digital library will provide even more redundancy to the users who fail to appreciate the quality of independent information sources.

4. Users are easily persuaded by a piece of information they expect to be true, even though its source may be totally unreliable; For example, a head of a state does not expect the attack from a neighboring country. Consequently, he will trust the spy reports, stating that accumulated foreign military forces across the border are just another army exercise. A poet (Ogden Nash) in one of his humorous pieces of wit, explains how the door on fanatic's mind opens outwardly, therefore the pressure of facts can just close it more firmly. As far as the digital library users are concerned, they should be taught how to get rid of any such "fanaticism". On the contrary, they must be open for the different information sources.

5. Complex human systems in most cases behave counter-intuitively. Consequently, whenever they follow intuition, the (digital information) users tend to reject correct models which are not supported by their own experience or expectations; For example, a medieval scientist rejects the idea that the Earth's shape might be spherical. It is because his own experience and belief are supporting the idea of the Earth being flat.

   The digital library of the future, as I see it, will become a plug-in, virtual, truly multicultural, content-based, just-in-time knowledge providing amusement park. It will serve as a point of departure into the virtual world of information, will be aimed at delighting the client,and will function as an expert system, a major catalyst of change for its environment. Without appropriate education, organization and standards, the users of digital libraries will be more often tempted by the abundance of unstructured and unorganized sources of information to make the above mentioned errors.

[ Return to Digital Libraries Workshop ]