Established Communities of Practice (CoP) can enhance the learning experience, especially for online learners, who often have an insufficient support structure when compared to on-campus learners. This is even more of a problem when the mode of instruction is asynchronous. To remedy this situation, we have built a framework for the establishment of Communities of Practice in virtual worlds, even when the majority of the members of the community are nonhuman intelligent agents. We refer to these groups as Synthetic Communities of Practice (SCoP). This paper details our current work, illustrates a software architecture to support SCoP formation, and describes results to date. We believe that this infrastructure will become important, especially with the emergence of Massive Open Online Courses (MOOC).

E-learning has attracted a great deal of interest in educational circles from K-12 to universities. A question that is often rightly asked is how effective current e-learning systems are. It is argued that there is little individualization of instruction by adapting to the pedagogical needs of each learner in current e-learning systems. Intelligent tutoring systems have tried to fill this gap but even they fail to compete with human one-to-one tutoring. This paper presents Affective Tutoring Systems which are e-learning systems capable of detecting learners’ affective state and reacting to it through a life like agent called Eve. This paper presents an Affective Tutoring System in the domain of mathematics and the research that led to its development. It also presents the findings from the study and testing of the system indicating that the animated agent Eve carried a persona effect.

This paper examines the effective deployment of conversational agents in virtual worlds from the perspective of researchers/practitioners in cognitive psychology, computing science, learning technologies and engineering. From a cognitive perspective, the major challenge lies in the coordination and management of the various channels of information associated with conversation/communication and integrating this information with the virtual space of the environment and the belief space of the user. From computing science, the requirements include conversational competency, use of nonverbal cues, animation consistent with affective states, believability, domain competency and user adaptability. From a learning technologies perspective, the challenge is to maximise the considerable affordances provided by conversational avatars in virtual worlds balanced against ecologically valid investigations regarding utility. Finally, the engineering perspective focuses on the technical competency required to implement effective and functional agents, and the associated costs to enable student access. Taken together, the four perspectives draw attention to the quality of the agent–user interaction, how theory, practice and research are closely intertwined, and the multidisciplinary nature of this area with opportunities for cross fertilisation and collaboration.

This paper examines the effective deployment of conversational agents in virtual worlds from the perspective of researchers/practitioners in cognitive psychology, computing science, learning technologies and engineering. From a cognitive perspective, the major challenge lies in the coordination and management of the various channels of information associated with conversation/communication and integrating this information with the virtual space of the environment and the belief space of the user. From computing science, the requirements include conversational competency, use of nonverbal cues, animation consistent with affective states, believability, domain competency and user adaptability. From a learning technologies perspective, the challenge is to maximise the considerable affordances provided by conversational avatars in virtual worlds balanced against ecologically valid investigations regarding utility. Finally, the engineering perspective focuses on the technical competency required to implement effective and functional agents, and the associated costs to enable student access. Taken together, the four perspectives draw attention to the quality of the agent–user interaction, how theory, practice and research are closely intertwined, and the multidisciplinary nature of this area with opportunities for cross fertilisation and collaboration. [ABSTRACT FROM AUTHOR]

In a survey of 62 enterprises in New Zealand, including the six major universities, we were interested in finding out the state of industrial and educational practice with respect to the field of user interface design for the web. Our research revealed that usability issues seem to have taken a backseat to other kinds of development concerns. There is general lack of formal education, knowledge and skills in usability methods, processes and techniques amongst designers and developers. We also found that most universities in New Zealand offer one or two Human-Computer Interaction (HCI) courses within their information technology undergraduate degree programmes as electives. Conversely, our research shows that elsewhere internationally, HCI has become a major area of study especially where usability is a major industrial concern. We discuss the problem, its implications and possible remedies.

This article introduces a novel technique for automatic selection of long Haar-like features for obiect detection. The evaluation of a long Haar feature is based on the Viola-Jones algorithm for object detection, in classifying the positive and the negative training samples. The application of a genetic algorithm in this research for locating the best feature space is possibly the best approach to this problem considering the size of the feature. In this context, each of the features which can better distinguish the positive and the negative samples is given a higher rank in the population and will survive one more generation. Applying the top 10 chromosomes for face detection was promising, which suggests that a small number of long Haar-like features can potentially be applied for rigid object detection.

Vision-based gesture recognition systems require identifying and tracking the boundaries of skin segments. In this article, we present four variations of the Mean-Shift algorithm with the applications of hand and face tracking in user interface applications. The first variation is based on continuous sampling of the boundaries of the kernel and changing the size of the kernel. The second variation addresses tracking of multiple skin blobs in the image sequence. The third algorithm provides a high level of tracking which we call a "Macro tracker" for tracking activity from multiple trackers. The fourth variation addresses the application of depth information in blob tracking. This additional variable makes the algorithm more robust in handling occlusion, which often occurs in user interface applications.

Face and hand tracking are important areas of research, related to adaptive human-computer interfaces, and affective computing. In this article we have introduced two new methods for boundary detection of the human face in video sequences: (1) edge density thresholding, and (2) fuzzy edge density. We have analyzed these algorithms based on two main factors: convergence speed and stability against white noise. The results show that “fuzzy edge density” method has an acceptable convergence speed and significant robustness against noise. Based on the results we believe that this method of boundary detection together with the mean-shift and its variants like cam-shift algorithm, can achieve fast and robust tracking of the face in noisy environment, that makes it a good candidate for use with cheap cameras and real-world applications.

Many software systems would significantly improve performance if they could interpret the nonverbal cues in their user's interactions as humans normally do. Currently, Intelligent Tutoring Systems (ITSs) (and other software systems) are unable to use nonverbal cues to interpret student's responses to instructional material as can human tutors. We believe that this capability is essential to adapt teaching strategy to the needs of the learner. An experiment was performed aimed at identifying what kinds of gestures are being used by students in a human-to-human learning context. We have identified a range of gestures being used in one-to-one tutoring environments and a dependency of gesture use on students' skill level. As a result, we suggest how the student model in an ITS should reflect this dependency. These results are applicable to HCI in general.

[13] E M Gray and W L Smith. On the limitations of software process assessment and the recognition of a required reorientation for global process improvement.

In a pilot study, professional information technologists from both commercial and academic institutions showed that representing the requirements for a simple television remote control device could be done accurately and completely using separate languages that are for 1) user interface specification, and 2) system specification, and, 3) with natural language. That individuals can create 3 separate, but equivalent representations is important because, with the aid of an underlying meta-representation, it allows system stakeholders to view software requirements in whatever way best represents their particular area of concern, while maintaining internal consistency between views.

This paper reports on the progress made on the development of a Haskell tutor called the next generation intelligent tutoring system (NGITS). The project is aimed at developing a more human like intelligent tutoring system. The tutor being developed takes the learner's internal or emotional state into consideration when modelling the learner. It uses a camera to capture facial expressions and adds this information to the student's knowledge state which is developed using a case based reasoning approach. It intends to use other biometric data for the same purpose. The tutor will employ multiple teaching strategies and will switch between them based on the facial expressions and other information contained in, the student model.

Intelligent tutoring systems (ITS) provide individualised instruction. They offer many advantages over the traditional classroom scenario: they are always available, non-judgemental and provide tailored feedback resulting in increased and effective learning. However, they are still not as effective as one-on-one human tutoring. The next generation of intelligent tutors are expected to be able to take into account the cognitive and emotional state of students. This paper reports on the progress made in the development of a facial expression analysis component for intelligent tutoring systems. A digital camera is used to grab the image of the learner and signal processing techniques such as Wavelet and ANN are employed to extract facial expressions. We have developed algorithms for detecting learner’s face and locating permanent facial features such as eyebrow, eyes, and mouth. Facial expression analysis is performed based on both permanent and transient facial features in a nearly frontal-view face image sequence. This information will be added to the student’s knowledge state model. Other non-verbal interactions like heartbeat and eye and body movement will be used in the future to enhance the performance of the system. This will enable intelligent tutors to react to changes in student’s state reflected in student models. We have called this new generation of intelligent tutors, “affective tutoring systems”.

Intelligent tutoring systems (ITS) provide individualized instruction. They offer many advantages over the traditional classroom scenario: they are always available, nonjudgmental and provide tailored feedback resulting in increased and effective learning. However, they are still not as effective as one-on-one human tutoring. The next generation of intelligent tutors is expected to be able to take into account the cognitive and emotional state of students. We present a proposed contribution of affect to student modeling, and reports on the progress made in the development of a facial expression analysis component for intelligent tutoring systems.

People who use multiple channels at the same time communicate more successfully about spatial problems than those who rely exclusively on either voice or pictures. To achieve a similarly successful interaction between a person and a geographic information system (GIS), we use two concurrent communication channels—graphics and speech—to construct a multi-modal spatial query language in which users interact with a geographic database by drawing sketches of the desired configuration, while simultaneously talking about the spatial objects and the spatial relations drawn. Through the combined use of graphics and sketch, more intuitive and more precise specifications of spatial queries are possible. The key to this interaction is the exploitation of complementary or redundant information present in both graphical and verbal descriptions of the same spatial scenes. A multiple-resolution model of spatial relations is used to capture the essential aspects of a sketch and its corresponding verbal description. The model stresses topological properties, such as containment and neighborhood, and considers metrical properties, such as distances and directions, as refinements where necessary. This model enables the retrieval of similar, not only exact, matches between a spatial query and a geographic database. Such new methods of multi-modal spatial querying and spatial similarity retrieval will empower experts as well as novice users to perform easier spatial searches, ultimately providing new user communities access to spatial databases. 1 . Introduction Today’s methods of interacting with geographic databases are largely non-spatial, as they require their users to deal with geographic data primarily through alphanumeric command languages. Currently, spatial querying is done by typing a command in some spatial query language, such as an extended version of SQL (Ingram and Phillips 1987; Herring et al. 1988; Egenhofer 1994), or by selecting the same or a similar syntax through a forms interface or from pull-down menus (Egenhofer 1990; Calcinelli and Mainguenaud 1994; Aufaure-Portier 1995). Such spatial querying is a tedious process, because it often requires extensive training in the use of the particular query language. A more serious disadvantage of such textual spatial querying is that it forces users to translate a spatial image they may have in their minds about the situation they are interested in, into a non-spatial language. Graphical user interfaces provide only little improvement for such query languages, because they use the same type of syntax and grammar as the typed languages, and they only release users from remembering the particular syntax (Egenhofer 1992). The problems with communicating a user’s request to a spatial database through conventional spatial query languages become most apparent when several users have to work together and have to understand their intentions. Verbal descriptions of spatial situations are frequently ambiguous and may easily lead to misinterpretations, particularly in multi-language working groups. Traditional spatial query languages have serious limitations when geographic concepts are used that are vague, imprecise, little understood, or not standardized. As an example, take the notion of the spatial predicate “cross” whose semantics may vary depending on the context in which it is used, the meaning of * This work was partially supported by Rome Laboratories under grant number F30602-95-1-0042. Max Egenhofer’s research is further supported by grants from the National Science Foundation under grant number No. SES 88-10917 for the NCGIA and grant number IRI-9309230; the Scientific and Environmental Division of the North Atlantic Treaty Organization; Intergraph Corporation; Environmental Systems Research Institute Inc.; Space Imaging, Inc., and by a Massive Digital Data Systems contract sponsored by the Advanced Research and Development Committee of the Community Management Staff and administered by the Office of Research and Development.

Project managers can make more effective and efficient project adjustments if they detect project high-risk elements early. We analyzed 42 software development projects in order to investigate some early risk factors and their effect on software project success. Developers in our organization found the most important factors for project success to be: (1) the presence of a committed sponsor and (2) the level of confidence that the customers and users have in the project manager and development team. However, several other software project factors, which are generally recognized as important, were not considered important by our respondents.

Despite the advantages that object technology can provide to the software development community and its customers, the fundamental problems associated with identifying objects, their attributes, and methods remain: it is a largely manual process driven by heuristics that analysts acquire through experience. While a number of methods exist for requirements development and specification, very few tools exist to assist analysts in making the transition from textual descriptions to other notations for object-oriented analysis and other conceptual models. In this paper we describe a methodology and a prototype tool, Linguistic assistant for Domain Analysis (LIDA), which provide linguistic assistance in the model development process. We first present our methodology to conceptual modeling through linguistic analysis. We give an overview of LIDA's functionality and present its technical design and the functionality of its components. We also provide a comparison of LIDA's functionality with that of other research prototypes. Finally, we present an example of how LIDA is used in a conceptual modeling task.

Abstract : LIDA (Linguistic Assistant for Domain Analysis) helps analysts to develop object-oriented models of a domain, using a subset of UML. In order to develop such models, the requirements analyst or knowledge engineer often needs to analyze large volumes of text from "legacy documents" these might include user manuals of legacy systems, company policies, use cases, or transcripts of interviews with domain experts. LIDA facilitates this analysis by compiling a list of the words and multi-word terms in a document, and providing a graphical interface for the user to mark them as corresponding to elements of a model. It also lets the user validate models as they are created, through integration with CoGenTex's ModelExplainer tool, which generates textual descriptions of a model.