Ye-Sho Chen

A relationship between Lotka's Law, Bradford's Law, and Zipf's Law

A common functional relationship among Lotkas law, Bradfords law, and Zipfs law is derived. The proof takes explicit account of the sequences of observed values of the variables by means of an index. This approach results in a more realistic and precise formulation of each law.

Mathematical modeling of empirical laws in computer applications: A case study

Abstract A major difficulty in using empirical laws in computer applications is the estimation of parameters. In this paper, we argue that the difficulty arises from the misuse of goodness-of-fit tests. As an alternative, we suggest the use of Simons theory of model building and apply the theory to examine the two well-known laws of Zipf. As a result, we show that the Simon-Yule model of text generation derives two formulations of Zipfs law: (1) the frequency-count distribution proposed by Booth, and (2) the frequency-rank distribution proposed by Mandelbrot. A further significant contribution of the paper is that it provides a theoretical foundation for the estimation of parameters associated with the two laws of Zipf.

Mathematical and computer modelling of the Pareto principle

The well-known Pareto principle states that approximately 80% of wealth was concentrated in about 20% of a population. Two problems have been reported on the abuse of the 80/20 phenomenon: (1) the measure deviates from 80/20 in many applications of the principle; under what condition, then, will the 80/20 rule be true? And, (2) if the 80/20 measure is correct at the time of analysis, will it be true over time? In this paper, we address these two problems by conducting mathematical modelling and computer simulations of the Pareto principle. Several significant factors influencing the phenomenon are identified.

A pictorial approach to poor-quality cost management

A pictorial approach to modeling poor-quality cost (PQC) is proposed. The approach is patterned after that used in a computer-based information system design. It includes two major steps: (1) specifying the PQC variables as well as the significant relationship among the variables, and mapping the variables and relationships into an influence diagram showing the structure of a PQC system: and (2) converting the structure into a well-defined entity-relationship diagram showing the input-output functions and their associated properties. Two major benefits of using the pictorial approach are that the influence diagram can provide an easy-to-understand PQC system for quality management practitioners and the entity-relationship diagram can provide an effective framework for maintaining and revising the PQC system. >

THEORETICAL FOUNDATION OF THE 80/20 RULE

A rigorous analysis of the 80/20 rule is made using an index for the observed values of the variables. Three important findings are identified. First, a sufficient condition is provided for Burrells inverse relationship between minimum holdings and average circulation rate. Second, an indexed version of Lotkas law is used to derive a sufficient condition for Egghes finding on the 80/20 rule. Third, through the computer simulations of the Simon-Yule model of Lotkas law, we identify the entry rate of new holdings as well as the number of circulations when the entry rate is a decreasing function to be crucial factors for the pattern of the 80/20-type curve.

Cyber security management and e-government

With the rapid growth of Information Technology, organisations are taking extra precautions when it comes to protecting information. In the complex interconnected business world, the common practice of deploying company-wide security solutions (e.g., anti-virus software and firewalls) is only the very first step of cyber security management. Adding to this first layer of security defence, two business-oriented security solutions are included in this paper: global partnership and proactive participation. In addition, an evolutionary process of learning to enhance the three layers of cyber security management is recommended to be done continuously. In order to manage these three layers of cyber security economically and effectively, an intranet-based framework of cyber security knowledge repositories based on the 80/20 rule is also proposed

The Simon-Yule approach to bibliometric modeling

Abstract Using an index approach to take into account the scattering pattern of the observed values, Chen and Leimkuhler showed that the three well-known bibliometric distributions (i.e., Lotkas law of scientific productivity, Bradfords law of bibliographic scattering, and Zipfs law of word frequency) are equivalent. Furthermore, Chen showed that Lotkas law can be derived from a generating mechanism (the Simon-Yule Model) proposed by Herbert A. Simon. In this paper, we use a simulation algorithm based on the Simon-Yule model to conduct computational experimentation on these three laws. The results indicate that the probability of a new entry (α), be it constant or decreasing, determines the characteristics of all three distributions.

Analysis of Lotka's law: The Simon-Yule approach

Abstract A major difficulty in using the well-known Lotkas law in information science is in the estimation of parameters. In this paper, we argue that the difficulty arises from the misuse of goodness-of-fit tests. As an alternative, we adopt Simons five-step modeling process for the study of Lotkas law. Three significant contributions can be identified. First, an index approach is used to identify a general formulation of Lotkas law. Second, a time series approach is used to identify two influential variables associated with the empirical data. Third, the constructive mechanism proposed by Simon is used to derive a distribution resembling the general formulation of Lotkas law. Further research on refining the constructive mechanism is suggested.

Bradford's law: An index approach

A rigorous analysis of Bradfords law is made using an index for the observed values of the variables. Three important properties relating size and frequency are identified. Using this approach, the shape of Bradford-type curves can be described in terms of three distinct regions and two shape parameters.

An entity-relationship approach to design support and expert systems

Abstract There has been an increasing interest in integrating decision support systems and expert systems to provide decision makers a more accessible, productive and domain-independent information and computing environment. In this paper, we review a workstation-based expert decision support system (WXDSS) proposed by Chen and Pruett. A database-oriented design process is discussed consisting of four phases: (1) requirements analysis, (2) conceptual framework, (3) logical design, and (4) physical system implementation. An entity-relationship (ER) approach to the design of conceptual framework is studied. A positive characteristic of the ER approach is that it provides the user an enterprise view of the WXDSS that is independent of how the information is stored and processed.