Towards Next Generation Web: Knowledge Web

The Right Needle from the Right Haystack

Among the technological generations of World Wide Web, content of any kind and for any audience grew and grew. Meanwhile the user is overwhelmed by a flood of information offered to him indiscriminately by the World Wide Web, which leads to a significantly decreasing usability. Search engines have not (yet) improved this picture. They introduce their own policies and biases. They filter out useful information and direct the searcher in a specific way. Search is moreover still bound to addresses which may be known or unknown to the searcher.

The user is typically not able to determine the quality of data obtained. He or she can

• •

  neither judge accuracy, currency, appropriateness, completeness, correctness, independence, learnability, maturity, reliability, stability, suitability, or understandability,

• •

  nor decide whether data obtained are at the right moment of time, of the right kind, in the right dose, of the right form, in the complete extent, and within the restrictions agreed upon in advance.

There is no authority for quality. The business model for the World Wide Web is based on freedom of content to a certain extent. This freedom may be nicely used or may be misused. Wikipedia is a typical example of partially very useful and partially completely confusing information.¹

Missing Adaptivity and User Orientation

Information as processed by humans is perceived in a very subjective way. For a web information system, the determining factor as to whether the user can derive advantage from the content delivered is the user's individual situation (Kobsa, 2005), i.e., the life case, user model and context. The same category of information can cause various needs in different life cases. For instance, a divorcee has a completely different need for information in fatherhood than a prospective father, although both of them reside in the same category. It is not the case that any user can deal with any kind of content. For the casual user or the novice other content has to be delivered than for experts. The common web information system doesn't reflect the user's situation and neglects the user's specific needs. As a result, the user is spammed with information which is predominantly out of focus. The abundance of information also makes it impossible for the user to separate useful from useless content. By the absence of metadata, any unspecified information reduces the usability of the World Wide Web as a whole. Whether the content obtained from the web may be classified as information depends on the user. There is no commonly agreed definition of information. We may define information as follows within the five layer model of (Mortiz, Schewe, Thalheim, 2005)(Murphy, 2001)(Schewe and Thalheim, 2007):

Information, as processed by human users of web sites, is

• •

  data

• •

  perceived or noticed, selected and organized by its receiver,

• •

  because of his subjective human interests, originating from his instincts, feelings, experience, intuition, common sense, values, beliefs, personal knowledge, or wisdom,

• •

  simultaneously processed by his cognitive and mental processes and

• •

  seamlessly integrated in his recallable knowledge.

Therefore, web site modeling must include a description of the information need of actors. The information need can be specified as

• •

  conceptual incongruity in which a person's cognitive structure is not adequate to a task,

• •

  when a person recognizes that something is wrong in their state of knowledge and desires to resolve the anomaly,

• •

  when the current state of possessed knowledge is less than is needed,

• •

  when internal sense runs out and

• •

  when there is insufficient knowledge to cope with voids, uncertainty or conflict in a knowledge area.

Moreover, users are limited

• •

  in their abilities for verbalization,

• •

  in their abilities for digestion of data and

• •

  by their habits, practices and cultural environment.

These limitations may cause intellectual overburdening of users. Most systems that require sophisticated learning courses for their exploration and utilization did not consider these limitations and did not cope with real life situations. The approach we use for avoiding overload is based on observation of real applications before developing the web information system.