Design Consideration of Sociomaterial Multi-Agent CSCW Systems

The recent technological advancements have significantly redefined the context in which organizations do business processes including the processes used to acquire, process, and share information. The transformations that emerged across the organizational and institutional landscapes have led to the emergence of new organizational forms of design and new business models. Within this context, the new business patterns, platforms, and architectures have been developed to enable for the maximization of benefits from data through the adoption of collaborative work practices. The main focus of such practices is oriented towards the improvement of responsiveness, building of alliances, and enhancing organizational reach. The use of global networks and Web-based systems for the implementation of collaborative work has been accompanied with a wide range of computer-supported collaborative systems. This chapter examines the context of collaboration, collaborative work, and the development of agent-supported collaborative work system. It also examines the implications of the ontological positions of sociomateriality on agent-supported collaborative work domains in terms of the multi-agent architecture and multi-agent evaluation.

Risk Prediction Model for Osteoporosis Disease Based on a Reduced Set of Factors

The health industry collects huge amounts of health data, which, unfortunately, are not mined to discover hidden information. Information technologies can provide alternative approaches to the diagnosis of the osteoporosis disease. In this chapter, the authors examine the potential use of classification techniques on a huge volume of healthcare data, particularly in anticipation of patients who may have osteoporosis disease through a set of potential risk factors. An innovative solution approach based on dynamic reduced sets of risk factors using the promising Rough Set theory is proposed. An experimentation of several classification techniques have been performed leading to rank the suitable techniques. The reduction of potential risk factors contributes to enumerate dynamically optimal subsets of the potential risk factors of high interest leading to reduce the complexity of the classification problems. The performance of the model is analyzed and evaluated based on a set of benchmark techniques.

Influencing Actions-Related Decisions Using Soft Computing Approaches

This chapter describes soft computing approaches for human-agent communications in the context of influencing decision-making behavior for health-related actions. Several methods are illustrated including using a person's predispositions and generalization techniques that allow issues to be viewed in a more favorable light with social interaction persuasion tendencies modeled with soft computing. The context of a robotic assistant for the elderly is used to illustrate the various communication techniques. Hierarchical generalization is introduced as a technique for generating potential alternatives in choices that might be more broadly acceptable to an individual who is being motivated towards a better choice. Finally, the related topic of negotiations using some the developed techniques is presented.

Segmentation of Renal Calculi in Ultrasound Kidney Images Using Modified Watershed Method

US images are a commonly used tool for renal calculi diagnosis, although they are time consuming and tedious for radiologists to manually detect and calculate the size of the renal calculi. It is very difficult to properly segment the US image to detect interested area of objects with the correct position and shape due to speckle formation and other artifacts. In addition, boundary edges may be missing or weak and usually incomplete at some places. With that point of view, the proposed method is developed for renal calculi segmentation. A new segmentation method is proposed in this chapter. Here, new region indicators and new modified watershed transformation are utilized. The proposed method is comprised of four major processes, namely preprocessing, determination of outer and inner region indictors, and modified watershed segmentation with ANFIS performance. The results show the effectiveness of proposed segmentation methods in segmenting the kidney stones and the achieved improvement in sensitivity and specificity measures.

Welcome to the Party

Modeling topic distributions over documents has become a recent method for coping with the problematic of huge amounts of unstructured data. Especially in the context of Web communities, topic models can capture the zeitgeist as a snapshot of people's communication. However, the problem that arises from that static snapshot is that it fails to capture the dynamics of a community. To cope with this problem, dynamic topic models were introduced. This chapter makes use of those topic models in order to capture dynamics in user behavior within microblog communities such as Twitter. However, only applying topic models yields no interpretable results, so a method is proposed that compares different political parties over time using regression models based on DTM output. For evaluation purposes, a Twitter data set divided into different political communities is analyzed and results and findings are presented.

Determinants for the Goodness of Performance Measurement Systems

The supply of adequate information is one of the main functions of Performance Measurement Systems (PMS), but also one of its drawbacks and reason for failure. Not only the collection of indicators is crucial, but also the stakeholders' understanding of their meaning, purpose, and contextual embedding. Today, companies seek a PMS without a way to express the goodness of a solution, indicating its ability to deliver appropriate information and to address these demands. The goal of this chapter is to explore the mechanisms that drive information and knowledge supply in PMS in order to model a way to express this goodness. Using a grounded theory approach, a theory of visibility of performance is developed, featuring a catalog of determinants for the goodness of PMS. Companies can conveniently use them to assess their PMS and to improve the visibility of their performance.

A Smart and Dynamic Decision Support System for Nonlinear Environments

The design of a dynamic and efficient decision-making system for real-world systems is an essential but challenging task since they are nonlinear, chaotic, and high dimensional in nature. Hence, a Support Vector Machine (SVM)-based model is proposed to predict the future event of nonlinear time series environments. This model is a non-parametric model that uses the inherent structure of the data for forecasting. The dimensionality of the data is reduced besides controlling noise as the first preprocessing step using the Hybrid Dimensionality Reduction (HDR) and Extended Hybrid Dimensionality Reduction (EHDR) nonlinear time series representation techniques. It is also used for subsequencing the nonlinear time series data. The proposed SVM-based model using EHDR is compared with the models using Symbolic Aggregate approXimation (SAX), HDR, SVM using Kernel Principal Component Analysis (KPCA), and SVM using varying tube size values for historical data on different financial instruments. A comparison of the experimental results of the proposed model with other models taken for the experimentation has proven that the prediction accuracy of the proposed model is outstanding.

Evaluation of Topic Models as a Preprocessing Engine for the Knowledge Discovery in Twitter Datasets

In the World Wide Web, users are an important information source for companies or institutions. People use the communication platforms of Web 2.0, for example Twitter, in order to express their sentiments of products, politics, society, or even private situations. In 2014, the Twitter users worldwide submitted 582 million messages (tweets) per day. To process the mass of Web 2.0's data (e.g. Twitter data) is a key functionality in modern IT landscapes of companies or institutions, because sentiments of users can be very valuable for the development of products, the enhancement of marketing strategies, or the prediction of political elections. This chapter's aim is to provide a framework for extracting, preprocessing, and analyzing customer sentiments in Twitter in all different areas.

Strategies and Methods for Ontology Alignment

The concept of ontologies has been around for millennia and spans many domains and disciplines. Ontologies are a powerful concept when applied to intelligent computing. Ontologies are the backbone of intelligent computing on the World Wide Web and crucial in many decision-support situations. Many sophisticated tools have been developed to support working with ontologies, including prominently exploiting the vast array of existing ontologies. Systems have been developed to automatically generate, match, and integrate ontologies in a process called ontology alignment. This chapter extends the current literature by presenting a system called ALIGN, which demonstrates how to use freely available tools to develop and facilitate ontology alignment. The first two ontologies are built with the ontology editor Protégé and represented in OWL. ALIGN then accesses these ontologies via Java's JENA framework and SPARQL queries. The efficacy of the ALIGN prototype is demonstrated on a drug-drug interaction problem. The prototype could readily be applied to other domains or be incorporated into decision-support tools.

Hybrid Privacy Preservation Technique Using Neural Networks

A novel Artificial Neural Network (ANN) dimension expansion-based framework that addresses the demand for privacy preservation of low dimensional data in clustering analysis is discussed. A hybrid approach that combines ANN with Linear Discriminant Analysis (LDA) is proposed to preserve the privacy of data in mining. This chapter describes a feasible technique for privacy preserving clustering with the objective of providing superior level of privacy protection without compromising the data utility and mining outcome. The suitability of these techniques for mining has been evaluated by performing clustering on transformed data and the performance of the proposed method is measured in terms of misclassification and privacy level percentage. The methods are further validated by comparing the results with traditional Geometrical Data Transformation Methods (GDTMs). The results arrived at are significant and promising.