Design And Implementation Of A Distributed Recruitment Management System

2.4.1    Measuring a Persons Knowledge
How can we measure whether a person knows something? To measure something means to assign a number to a characteristic (knowledge) of an object (a person) or event according to a set of rules. It is the set of rules by which the number is assigned that defines the meaning of the number. The currently used multiple-choice test or any other epistemetric method may be considered as a set of rules by which the numbers (scores) or measurements are produced and thus, knowledge may be operationally defined. Most tests used today for measuring a person’s knowledge in a field of study are aimed at composing test items that represent the field; and are fair and unbiased, (i.e. not influenced by the test takers’ characteristics other than knowledge, such characteristics include gender, ethnicity etc.) which might influence the measurement. To determine whether a person possesses knowledge on, say, simple addition, we can ask questions that are representative of the topic, such as
What is the sum of 12 + 13? OR we might pose the question as a response selection or multiple-choice task, e.g.
12+13 = (a) 7
              (b) 14
              (c) 24
              (d) 25
Current testing practice is to observe which alternative a person selects and infer that she/he knows (if a correct answer is selected) or does not know (if the correct answer is not selected) how to add two digit numbers. However, a test taker can select the correct answer without knowing how to add, e.g. in the above example, the chance of being correct by guessing alone is 1/4 = 25%. The reliance exclusively on the correctness of the answer implies that the person who provides a correct but unsure answer or who made a lucky guess possesses knowledge equivalent to a person who is correct and extremely sure of it.
Similarly, in today’s multiple-choice tests if an incorrect answer is selected, then it is interpreted simply to mean that the person does not know the answer, i.e. is uninformed. This inference is misleading. Specifically, the person may be extremely sure that the incorrect answer which he/she selected is correct and, thus, may be misinformed which is much worse than being uninformed. A sure-but-wrong belief, used confidently as a basis for making decisions and taking actions, may lead to surprising errors in performance sometimes with tragic results. For example, if a licensing or certification test is being administered to a professional (say a physician or an aspiring key decision maker), it is important to make the distinction for incorrect answers between a person who:
1. is not sure at all as to whether an incorrect answer which he/she gave is correct and thus the incorrect belief is not likely to be employed in practicing the profession; or
2.Strongly believes that the selected incorrect answer is correct and is therefore likely to use the erroneous belief in making decisions,
An easy way is to think of the person who makes a true statement based on adequate reasons, but does not feel confident that it is true. Obviously, he is much less likely to act on it, and, in the extreme case of lack of confidence, would not act on it (Pears, 1971).
2.4.2    Knowledge Discovery in Databases
Knowledge Discovery in Databases is the nontrivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns in data (Fayyad, 1996). Here, data are a set of facts (for example, cases in a database), and pattern is an expression in some language describing a subset of the data or a model applicable to the subset. Hence, in our usage here, extracting a pattern also designates fitting a model to data; finding structure from data; or, in general, making any high-level description of a set of data (Fayyad et al, 1996).
The term process implies that KDD comprises many steps, which involve data preparation, search for patterns, knowledge evaluation, and refinement, all repeated in multiple iterations. By nontrivial, we mean that some search or inference is involved; that is, it is not a straightforward computation of predefined quantities like computing the average value of a set of numbers (Fayyad et al, 1996).