Learning machines: Rationale and application in ground-level ozone prediction

Multilayer perceptron (MLP) and support vector machine (SVM), two popular learning machines, are increasingly being used as alternatives to classical statistical models for ground-level ozone prediction. However, employing learning machines without sufficient awareness about their limitations can lead to unsatisfactory results in modeling the ozone evolving mechanism, especially during ozone formation episodes. With the spirit of literature review and justification, this paper discusses, with respect to the concerning of ozone prediction, the recently developed algorithms/technologies for treating the most prominent model-performance-degradation limitations. MLP has the “black-box” property, i.e., it hardly provides physical explanation for the trained model, overfitting and local minima problems, and SVM has parameters identification and class imbalance problems. This commentary article aims to stress that the underlying philosophy of using learning machines is by no means as trivial as simply fitting models to the data because it causes difficulties, controversies or unresolved problems. This article also aims to serve as a reference point for further technical readings for experts in relevant fields.     

Design and evaluation of virtual environments mechanisms to support remote collaboration on complex process diagrams

Many organizational analysis tasks are solved by collaborating teams. In technology-mediated collaborations, enabling relevant visual cues is a core issue with existing technology. We explore whether avatars can provide relevant cues in collaborative virtual environments. To do so, we develop a research prototype for a collaborative virtual environment that utilises avatars to improve workspace awareness in collaborative tasks. We test this solution through two studies, qualitative and quantitative, in which participants have to collaborate to jointly validate and correct a diagrammatic model of operational procedures.Our evaluations provide both positive and negative results about the proposed prototype. Collaboration patterns changed and became significantly easier, but, task performance was only slightly improved. Together, these findings inform both the ongoing development of collaborative 3D virtual environments and the role of technology-mediated collaboration for validating and fixing models of processes.     

Knowledge-based support in help-desk environments

How much does a knowledge based system improve help desk interactions? To investigate the relationship between a KBS and a help desk and to provide a first step in determining the return on investment of KBSs in general, we conducted a study that showed as accurately as possible the impact of a KBS on one of Hewlett-Packard's customer support centers. We organized the study to verify three main assumptions: it takes less time to resolve a call when the analyst uses the KBS; the time saving shows up as a weekly decrease in overall interaction time; using a KBS allows the analyst to solve more kinds of problems than not using it     

Collaborative virtual learning environments: design and evaluation

E-learning systems have gone through a radical change from the initial text-based environments to more stimulating multimedia systems. Such systems are Collaborative Virtual Environments, which could be used in order to support collaborative e-learning scenarios. The main aim of this paper is to aid educational designers in selecting, designing and evaluating three dimensional collaborative virtual environments in order to gain the pedagogical benefits of Computer Supported Collaborative Learning. Therefore, this paper initially discusses the potential of three dimensional networked virtual environments for supporting collaborative learning. Furthermore, based on a two-step platform selection process this paper (a) presents and compares three dimensional multi-user virtual environments for supporting collaborative learning and (b) validates the most promising solution against a set of design principles for educational virtual environments. According to these principles, an educational environment has been implemented on top of the selected platform in order to support collaborative e-learning scenarios. The design of this environment is also presented. In addition, this paper presents the results of three small scale studies carried out in a tertiary education department, to assess the educational environment. This environment has been evaluated based on a hybrid evaluation methodology for uncovering usability problems, collecting further requirements for additional functionality to support collaborative virtual learning environments, and determining the appropriateness of different kinds of learning scenarios.

The Cactis project: database support for software environments

The Cactis project is an on-going effort oriented toward extending database support from traditional business-oriented applications to software environments. The main goals of the project are to construct an appropriate model, and develop new techniques to support the unusual data management needs of software environments, including program compilations, software configurations, load modules, project schedules, software versions, nested and long transactions, and program transformations. The ability to manage derived information is common to many of these data needs, and the Cactis database management system has the ability to represent and maintain derived data in a time- and space-efficient fashion. A central contribution of Cactis is its integration of the type constructors of semantic models and the localized behavior capabilities of object-oriented database management systems     

ParaGraph: Graph editor support for parallel programming environments

We report here on a graph editor, ParaGraph, that supports massively parallel programming. It provides a flexible mechanism for the concise specification of families of annotated graphs, addressing the problems of user annotation and scale independent graph manipulation. ParaGraph currently serves as the basis for tools supporting communication abstractions in program specification and debugging. Its foundation in an extended form of aggregate rewriting graph grammars makes its adaptation to other parallel programming environments straightforward.The Parallel Programming Environments Project at the University of Massachusetts is supported by the Office of Naval Research under Contract N000014-84-K-0647 and by the National Science Foundation under Grants DCR-8500332 and CCR-8712410.     

Learning patterns in ambient intelligence environments: a survey

It is essential for environments that aim at helping people in their daily life that they have some sort of Ambient Intelligence. Learning the preferences and habits of users then becomes an important step in allowing a system to provide such personalized services. Thus far, the exploration of these issues by the scientific community has not been extensive, but interest in the area is growing. Ambient Intelligence environments have special characteristics that have to be taken into account during the learning process. We identify these characteristics and use them to highlight the strengths and weaknesses of developments so far, providing direction to encourage further development in this specific area of Ambient Intelligence.     

Top management support in multiple-project environments: an in-practice view

Top management support has generally been considered the most critical factor for the success of IS projects. Typically, there have developed implicit or explicit assumptions that top management support has to be constant and consistent during the entire life of an IS implementation project. However, previous research investigating this issue has been based mainly on a ‘single project’ perspective, although contemporary organisational settings have increasingly involved many projects taking place simultaneously. Such multiple-project environments bring into question the feasibility of the previous assumptions and invite revisiting them. This paper aims to do this by examining the materialisation of top management support in a multiple-project environment. To this end, it investigates the interactions between a project to implement an Enterprise Resource Planning system and other projects and programmes in a prominent international organisation. It applies a framework from Actor Network Theory to explore the different aspects of what is involved in being in a multiple-project environment and the nature of the materialisation of top management support. The findings unravel some of the complexity surrounding top management support in this contemporary organisational setting. Some important implications for theory and practice are highlighted and discussed.     

Context-dependent awareness support in open collaboration environments

The widespread adoption of online services for performing work, home and leisure tasks enables users to operate in the ubiquitous environment provided by the Internet by managing a possibly high number of parallel (private and shared) activity contexts. The provision of awareness information is a key factor for keeping users up-to-date with what happens around them; e.g., with the operations performed by their collaborators. However, the delivery of notifications describing the occurred events can interrupt the users’ activities, with a possible disruptive effect on their emotional and attentional states. As a possible solution to the trade-off between informing and interrupting users, we defined two context-dependent notification management policies which support the selection of the notifications to be delivered on the basis of the user’s current activities, at different granularity levels: general collaboration context versus task carried out. These policies are offered by the COntext depeNdent awaReness informAtion Delivery (CONRAD) framework. We tested such policies with users by applying them in a collaboration environment that includes a set of largely used Web 2.0 services. The experiments show that our policies reduce the levels of workload on users while supporting an up-to-the-moment understanding of the interaction with their shared contexts. The present paper presents the CONRAD framework and the techniques underlying the proposed notification policies.     

Learning and planning in environments with delayed feedback

This work considers the problems of learning and planning in Markovian environments with constant observation and reward delays. We provide a hardness result for the general planning problem and positive results for several special cases with deterministic or otherwise constrained dynamics. We present an algorithm, Model Based Simulation, for planning in such environments and use model-based reinforcement learning to extend this approach to the learning setting in both finite and continuous environments. Empirical comparisons show this algorithm holds significant advantages over others for decision making in delayed-observation environments.     

Multiserver support for large-scale distributed virtual environments

CyberWalk is a distributed virtual walkthrough system that we have developed. It allows users at different geographical locations to share information and interact within a shared virtual environment (VE) via a local network or through the Internet. In this paper, we illustrate that as the number of users exploring the VE increases, the server will quickly become the bottleneck. To enable good performance, CyberWalk utilizes multiple servers and employs an adaptive region partitioning technique to dynamically partition the whole VE into regions. All objects within each region will be managed by one server. Under normal circumstances, when a viewer is exploring a region, the server of that region will be responsible for serving all requests from the viewer. When a viewer is crossing the boundary of two or more regions, the servers of all the regions involved will be serving requests from the viewer since the viewer might be able to view objects within all these regions. This is analogous to evaluating a database query using a parallel database server, which could improve the performance of serving a viewer's request tremendously. We evaluate the performance of this multiserver architecture of CyberWalk via a detail simulation model.     

User-centered design and evaluation of virtual environments

We present a structured, iterative methodology for user-centered design and evaluation of VE user interaction. We recommend performing (1) user task analysis followed by (2) expert guidelines-based evaluation, (3) formative user-centered evaluation, and finally (4) comparative evaluation. In this article we first give the motivation and background for our methodology, then we describe each technique in some detail. We applied these techniques to a real-world battlefield visualization VE. Finally, we evaluate why this approach provides a cost-effective strategy for assessing and iteratively improving user interaction in VEs     

Socially augmented argumentation tools: Rationale,design and evaluation of a debate dashboard

Collaborative Computer-Supported Argument Visualization (CCSAV) is a technical methodology that offers support for online collective deliberation over complex dilemmas. As compared with more traditional conversational technologies, like wikis and forums, CCSAV is designed to promote more critical thinking and evidence-based reasoning, by using representations that highlight conceptual relationships between contributions, and through computational analytics that assess the structural integrity of the network. However, to date, CCSAV tools have achieved adoption primarily in small-scale educational contexts, and only to a limited degree in real world applications. We hypothesise that by reifying conversations as logical maps to address the shortcomings of chronological streams, CCSAV tools underestimate the importance of participation and interaction in enhancing collaborative knowledge-building. We argue, therefore, that CCSAV platforms should be socially augmented in order to improve their mediation capability. Drawing on Clark and Brennan influential Common Ground theory, we designed a Debate Dashboard, which augmented a CCSAV tool with a set of widgets that deliver meta-information about participants and the interaction process. An empirical study simulating a moderately sized collective deliberation scenario provides evidence that this experimental version outperformed the control version on a range of indicators, including usability, mutual understanding, quality of perceived collaboration, and accuracy of individual decisions. No evidence was found that the addition of the Debate Dashboard impeded the quality of the argumentation or the richness of content.     

Learning gender with support faces

Nonlinear support vector machines (SVMs) are investigated for appearance-based gender classification with low-resolution "thumbnail" faces processed from 1,755 images from the FERET (FacE REcognition Technology) face database. The performance of SVMs (3.4% error) is shown to be superior to traditional pattern classifiers (linear, quadratic, Fisher linear discriminant, nearest-neighbor) as well as more modern techniques, such as radial basis function (RBF) classifiers and large ensemble-RBF networks. Furthermore, the difference in classification performance with low-resolution "thumbnails" (21×12 pixels) and the corresponding higher-resolution images (84×48 pixels) was found to be only 1%, thus demonstrating robustness and stability with respect to scale and the degree of facial detail     

COCOA: COnversation-based service COmposition in pervAsive computing environments with QoS support

Pervasive computing environments are populated with networked services, i.e., autonomous software entities, providing a number of functionalities. One of the most challenging objectives to be achieved within these environments is to assist users in realizing tasks that integrate on the fly functionalities of the networked services opportunely according to the current pervasive environment. Towards this purpose, we present COCOA, a solution for COnversation-based service COmposition in pervAsive computing environments with QoS support. COCOA provides COCOA-L, an OWL-S based language for the semantic, QoS-aware specification of services and tasks, which further allows the specification of services and tasks conversations. Moreover, COCOA provides two mechanisms: COCOA-SD for the QoS-aware semantic service discovery and COCOA-CI for the QoS-aware integration of service conversations towards the realization of the user task’s conversation. The distinctive feature of COCOA is the ability of integrating on the fly the conversations of networked services to realize the conversation of the user task, by further meeting the QoS requirements of user tasks. Thereby, COCOA allows the dynamic realization of user tasks according to the specifics of the pervasive computing environment in terms of available services and by enforcing valid service consumption.     

Information and modeling resources for decision support in global environments

Decision support through information and modeling resources is crucial for strategic decision making in a global environment. This paper describes how decision support systems (DSS) were developed, implemented, and utilized in an organization to evaluate strategic options for foreign direct investments in manufacturing facilities. The paper shows how the process of decision support was implemented and integrated into the company's long range strategic and annual profit planning processes. The approach to decision support can be adopted by other firms to aid international investment planning and operations.     

Computer-aided evaluation of the housing environments

This paper deals with the development and environmental variables involved in designing the housing layout form. It is considered that the factors taken into consideration and the crileria used in the course of decision-making in creating and reshaping the residential environments should be systematically tested in the light of the needs and social goals of the community. As the main potential of the computer lies in the process of analysis and appraisal the main objective of this study is to evaluate the quality and performance of the physical environment by means of utilizing computer techniques.