Informational Data Mining

The previous decades have witnessed major change within the Information Systems (IS) environment with a corresponding emphasis on the importance of specifying timely and accurate information strategies. Currently, there is an increasing interest in data mining and information systems optimization. Therefore, it makes data mining for optimization of information systems a new and growing research community. This chapter surveys the application of data mining to optimization of information systems. These systems have different data sources and accordingly different objectives for knowledge discovery. After the preprocessing stage, data mining techniques can be applied on the suitable data for the objective of the information systems. These techniques are prediction, classification, association rule mining, statistics and visualization, clustering and outlier detection.

Fuzzy Analytic Hierarchy Process and Its Application to E-Marketplace Selection

Making decisions is a part of daily life. The nature of decision-making includes multiple and usually conflicting criteria. Multi Criteria Decision-Making (MCDM) problems are handled under two main headings: Multi Attribute Decision Making (MADM) and Multi Objective Decision Making (MODM). Analytic Hierarchy Process (AHP) is a widely used multi-criteria decision making approach and has successfully been applied to many practical problems. Traditional AHP requires exact or crisp judgments (numbers). However, due to the complexity and uncertainty involved in real world decision problems, decision makers might be more reluctant to provide crisp judgments than fuzzy ones. Furthermore, even when people use the same words, individual judgments of events are invariably subjective, and the interpretations that they attach to the same words may differ. This is why fuzzy numbers and fuzzy sets have been introduced to characterize linguistic variables. Here, the authors overview the most known fuzzy AHP approaches and their application, and they present a case study to select an e-marketplace for a firm, which produces and sells electronic parts of computers in Turkey.

Sustainable Waste Management System and Reverse Logistic Network Design in Plastic Industry

The shortage of natural sources and the threat of the bad trend have forced the industries to find environmentally-friendly alternatives and ecological approaches in their production line. In some countries, regulations have been issued for industries about this subject. Due to these reasons and more of them, logistic firms have been forced to take into consideration decreasing material and energy consumption and minimizing waste production in planning their network designs. In practice, it might be necessary to simultaneously optimize more than one conflicting objective to obtain effective and realistic solutions. In this chapter, current logistics network design of a plastic industry in Turkey has been investigated and reverse logistics network design has been developed to minimize waste production and to achieve green production. This chapter presents a mathematical model which is a fuzzy goal programming model for imprecise goals for reverse logistic network design with multiple objectives in plastic sector. The considered objectives are to reduce cost in reverse logistics, to improve product quality, and to provide environmental benefits by minimizing waste production.

Conventional and Non-Conventional Data Modeling

The modeling of database applications involves deciding on how to represent the project in real-world objects. The data modeling process corresponds to a set of conceptual tools to describe data, its relationships, its semantics, and constraints of consistency. This process involves the steps related to the identification of requisites, conceptual modeling of data, conventional modeling, and non-conventional modeling of objects, and its relationships. In the conceptual modeling, where there is no need to specify the methods and data flow, objects and their relationships are defined. In conventional modeling, in the mapping of the conceptual model (Entity/Relationship) to the logical model (Relational) conversion rules are applied. However, there are non-conventional resources with the ability to create and use data types based on a grouping of other data types. The user-defined objects can be defined and used like any other data type. This chapter describes the process of mapping the relational model for the object-relational modeling, using a practical application in agricultural context, but it should be noted that the methodology is applicable to any area of knowledge.

Database Systems in Biology

The successful sequencing of the genoma of various species leads to a great amount of data that need to be managed and analyzed. With the increasing popularity of high-throughput sequencing technologies, such data require the design of flexible scalable, efficient algorithms and enterprise data structures to be manipulated by both biologists and computational scientists; this emerging scenario requires flexible, scalable, efficient algorithms and enterprise data structures. This chapter focuses on the design of large scale database-driven applications for genomic and proteomic data; it is largely believed that biological databases are similar to any standard database-drive application; however, a number of different and increasingly complex challenges arises. In particular, while standard databases are used just to manage information, in biology, they represent a main source for further computational analysis, which frequently focuses on the identification of relations and properties of a network of entities. The analysis starts from the first text-based storage approach and ends with new insights on object relational mapping for biological data.

IP Layer Client Puzzles

It is very common today that many business models are based on offering on-line services. Profitability and efficiency of this business model relies on a secure and undisturbed Internet infrastructure. Unfortunately, services offered on Internet infrastructure, being an Open and yet untrusted network, are very often targets of Denial-of-Service and Distributed Denial-of-Service attacks. These attacks are today a serious problem for on-line services offered by many business models. Preventing or minimizing DoS and DDoS is a difficult task which could serve to many on-line service offering business models to provide quality services to their clients. The main objective of this chapter is to present the Client Puzzle mechanism as a new method designed to defend business networks and their on-line services from these attacks. By using a client puzzle protocol on the IP level, the client is forced to solve a cryptographic puzzle before it can request any operation from a server, thus creating computational efforts and delays to illegitimate attackers and minimizing their attack effects on services. In this chapter, the authors show that chained puzzle protocol reduces the network and insfrastructure overhead because the servers do not have to generate puzzles on a per-packet basis. In addition, the chapter analyzes the effectiveness and some limitations of chained puzzles method with regards to minimizing DDoS attacks and outlines a general approach for addressing the identified limitations. At the last part, the authors propose a solution based on the general principle that under attack legitimate clients should be willing to experience some degradation in their performance in order to obtain the requested service.

Web-Based Services and Future Business Models

A paradigm shift is taking place today that provides a compelling value proposition for organizations and requires the adoption of new business models for the management of their core activities in a competitive way. The new emerging business models are related with open innovation, cloud computing approach, as well as social networking, which creates opportunities for firms to harvest the resources and knowledge that could be found outside the firm’s boundaries. However, in order for firms to grasp most of the emerging technologies, they should reconfigure their activities to tackle the challenge and opportunity presented by new innovations and technology trends. In this chapter, the authors demonstrate the changes that these new trends are witnessing for the business models of firms from a provider and user perspective.

A BPM Framework for NPD Process Knowledge Management

In today’s context of strong competition among organizations and rapid changes in business surroundings, the organizations really need to start thinking about improving their performance, especially in knowledge intensive processes such as New Product Development. Business Process Management and Knowledge Management can represent organization’s strategic resources to the extent in which they are viewed as a base of success or failure, but they need to be supported by synergic systems that allow shaping the context in which knowledge is created and where knowledge can be re-used. Managing the explicit definition of the NPD processes and its resources allows the regulation of reusable “process knowledge,” the achievement of standardization, the improvement of best practice reuse, the improvement of time/cost efficiency, and the support of workers in the retrieval of knowledge resources suitable to conduct the product development activities. Thus, the aim of the chapter is to study how to best support companies in the collection and organization of process knowledge in the domain of their new product development, and to present an NPD process knowledge management framework which, starting from BPM approaches and its related technologies, allows the building of the required knowledge for the product development process more effectively for users and stakeholders.

A Modeling Approach to Building Wireless Sensors

The gap between wireless sensor networks and application experts such as doctors, physicists, and biologists is slowly closing. Previous efforts have been made to bring the two together, but a design and implementation methodology for the lone user has never been proposed. In this chapter, a procedure is proposed based on the author’s experience building and programming a wireless humidity sensor for a greenhouse with only a small amount of previous programming experience. Various factors affecting the design and construction of sensor nodes are analysed and then applied in a practical manner in the project. The project ended prematurely due to hardware faults but reached a point that allows the continuation of the methodology in a theoretical fashion.

A Review on Applied Data Mining Techniques to Stock Market Prediction

A portfolio manager considers forecasting the asset prices and measurement of the market risk of an underlying asset. Financial institutions produce datasets to handle their problems by using data mining tools. Recently new technologies have been developed for tracking, collecting, and processing financial data. From a data analysis point of view, this chapter reviews the published articles based upon predictive data mining applications to stock market index. It is observed that hybrid models that combine data mining techniques or integrate an algorithm to a method work efficiently. Finally, the chapter provides likely directions of future researches.