HCI Bibliography Home | HCI Journals | About IJMMS | Journal Info | IJMMS Journal Volumes | Detailed Records | RefWorks | EndNote | Hide Abstracts
IJMMS Tables of Contents: 282930313233343536373839

International Journal of Man-Machine Studies 38

Editors:B. R. Gaines; D. R. Hill
Publisher:Academic Press
Standard No:ISSN 0020-7373; TA 167 A1 I5
Links:Table of Contents
  1. IJMMS 1993 Volume 38 Issue 1
  2. IJMMS 1993 Volume 38 Issue 2
  3. IJMMS 1993 Volume 38 Issue 3
  4. IJMMS 1993 Volume 38 Issue 4
  5. IJMMS 1993 Volume 38 Issue 5
  6. IJMMS 1993 Volume 38 Issue 6

IJMMS 1993 Volume 38 Issue 1

Editorial BIB 1-2
  E. A. Edmonds
A Co-Operative Computer Based on the Principles of Human Co-Operation BIBA 3-22
  A. A. Clarke; M. G. G. Smyth
Co-operation is presented as a technique for radically improving human-computer interaction with complex knowledge bases during problem-identifying and problem-solving tasks. A study of human-human co-operation literature indicated the importance of creating an environment where the refinement of solutions can be based on argument and the resolution of differing viewpoints, as it is through this interaction that the nature of the problem is revealed. To bring about such an environment, the work identified and created three mechanisms now considered to be central to human-computer co-operation; goal-oriented working (GOW), an agreed definition knowledge base (ADKB), and a model which, using problem-domain rules, stimulates the interaction between the user and the machine: the partner model (PM). To identify the requirements of the co-operative machine more completely, a software exemplar was constructed, using the task metaphor of spatial design. The result of the work is the implementation of a machine software architecture which demonstrates the functioning of co-operation. This co-operative computer, its evaluators believe, supports a user-machine interaction having a totally new and different quality. The machine architecture and software tools and techniques developed in the work can form the foundation for building future co-operative systems.
Visualizing and Representing Knowledge for the End User: A Review BIBA 23-49
  B. S. Murray; E. McDaid
The field of visualization and its various aspects are discussed. Visual programming and program visualization are discussed, with particular reference to visual programming languages. A collection of recent systems taken from the scientific literature are then described. The factors of particular importance to the development of such systems are briefly mentioned. The outstanding areas for research are discussed with the finding of suitable graphical representations being seen as particularly important.
Constructing End-User Knowledge Manipulation Systems BIBA 51-70
  E. A. Edmonds; S. M. O'Brien; T. Bayley; E. McDaid
The automation of knowledge acquisition for expert and decision support systems is concerned primarily with the development of more powerful tools and environments for knowledge engineers. We propose that the acquisition process might benefit from a more direct participation by the domain expert. An architecture and software toolkit, which supports the construction of end-user knowledge manipulation systems (EUKMS), is presented. EUKMS are strongly committed to an external knowledge representation at the interface which uses domain-familiar abstractions, language and objects, and maps them to an internal, computational representation which is hidden from the user. If given a predefined domain model, the implementation of elegant graphical user interfaces to knowledge bases, automatic operationalization of a user's rules as code, plus facilities to modify the rules, represent a significant advance in the development of systems which enable non-programming personnel to create and refine rules, without the mediation of a knowledge engineer.
   An application of the toolkit in the creation of an interface to a speech knowledge system is discussed. Our experience suggests that (i) EUKMS are achievable, (ii) may have a useful role in an integrated environment of complementary tools and techniques for knowledge acquisition, and (iii) the external knowledge representation employed in the user interface contributes to direct manipulation by the end-user and enhances computer-human communication.
Knowledge-Based Systems in Speech Recognition: A Survey BIBA 71-95
  Stella M. O'Brien
Endeavours in large vocabulary (1000 word) automatic speech recognition (ASR) systems for task-oriented dialogues in which the syntax, semantics and pragmatics are inherently limited enjoy success within their domain, but it may prove impractical to build larger systems based on the techniques which enable this.
   There is a consensus of opinion among researchers in ASR that the application of speech-specific knowledge would facilitate the design of more sophisticated stochastic models than those being attempted currently, and reduce the amount of training data necessary for an acceptable level of performance. There have been several attempts at designing purely knowledge-based ASR systems and, more recently, hybrid classifiers which combine stochastic techniques and speech knowledge: some of these systems are described in this survey. Whichever approach is adopted, there are common problems of knowledge representation and control, which may continue to impede the progress of ASR until knowledge-based techniques progress.
Special Features of Plosives in Connected-Speech Signals BIBA 97-127
  Stella M. O'Brien
It is increasingly argued that continuous speech recognition systems must integrate both knowledge-based and stochastic components to optimize performance. Spectrograms are a source of acoustic-phonetic information which can be interpreted with reference to an awareness of articulatory possibilities and some of this knowledge may be formalized in terms of rules. Spectrogram interpreters appear to use an internal hierarchy of cues which can be restructured and used as part of an integrated complex of cues for identification purposes. The Speech Knowledge Interface (SKI) is used to allow a domain expert to formalize such knowledge in a machine-robust and implementable manner. This paper reports experiments and results arising from the interactive use of SKI in the identification of the place of articulation of plosives.
End-User Manipulation of a Knowledge-Based System: A Study and an Expert's Practice BIBA 129-145
  L. Candy; S. M. O'Brien; E. A. Edmonds
The advent of end-user manipulation of knowledge-based systems (EUKMS) provides new opportunities for addressing the problems of encapsulating domain expertise. Interfaces which enable the expert, a professional and/or scientific practitioner, to create, refine and evaluate rules about the constituent elements of their knowledge provide a means of circumventing some of the current barriers to successful knowledge encapsulation. The critical feature of the design of such systems is the provision of facilities for the automatic conversion of the expert's rules into code. In a study of scientific work involving the capture of phonetics expertise in a knowledge-based system, key aspects of a speech scientist's working practice were identified. This paper discusses that use of the Speech Knowledge Interface system (SKI) in the context of investigations into the construction of an enhanced model of speech production for a speaker independent, continuous speech recognizer. Evidence that providing the expert with an appropriate interface to a knowledge-based system stimulates questions about existing knowledge and gives rise to new insights into the scope of the investigations, was found. Thus, the process of knowledge externalization, both of knowledge which was only partially realized and knowledge that was perceived as "new" by the expert, was facilitated by the interaction with the system.

IJMMS 1993 Volume 38 Issue 2

Interactive Inductive Learning BIBA 147-167
  Michael Hadjimichael; Anita Wasilewska
We propose an interactive probabilistic inductive learning model which defines a feedback relationship between the user and the learning program. We extend previously described learning algorithms to a conditional model previously described by the authors, and formulate our Conditional Probabilistic Learning Algorithm (CPLA), applying conditions as introduced by Wasilewska to a probabilistic version of the work of Wong and Wong. We propose the Condition Suggestion Algorithm (CSA) as a way to use the syntactic knowledge in the system to generalize the family of decision rules. We also examine the semantic knowledge of the system implied by the suggested conditions and analyse the effects of conditions on the system. CPLA/CSA has been implemented by the first author and was used to generate the examples presented.
Planning Interactive Explanations BIBA 169-199
  Alison Casey
Human verbal explanations are essentially interactive. If someone is giving a complex explanation, the hearer will be given the opportunity to indicate whether they are following as the explanation proceeds, and if necessary interrupt with clarification questions. These interactions allow the speaker to both clear up the hearer's immediate difficulties as they arise, and to update assumptions about their level of understanding. Better models of the hearer's level of understanding in turn allow the speaker to continue the explanation in a more appropriate manner, lessening the risk of continuing confusion.
   Despite its apparent importance, existing explanation and text generation systems fail to allow for this sort of interaction. Although some systems allow follow-up questions at the end of an explanation, they assume that a complete explanation has been planned and generated before such interactions are allowed. However, for complex explanations interactions with the user should take place as the explanation progresses, and should influence how that explanation continues.
   This paper describes the EDGE system, which is able to plan complex, extended explanations which allow such interactions with the user. The system can update assumptions about the user's knowledge on the basis of these interactions, and uses this information to influence the detailed further planning of the explanation. When the user appears confused, the system can attempt to fill in missing knowledge or to explain things another way.
Neither Novice Nor Expert: The Discretionary User of Software BIBA 201-229
  Radhika Santhanam; Susan Wiedenbeck
Most studies that examine users' interaction with computers focus either on novices or experts. Discretionary users of computers, such as lawyers, executives, administrators and professors have been ignored. Using techniques of process analysis, this study investigated the characteristics of discretionary users by observing their interaction with commercial word-processing software. Their performance was compared to groups of novices and experts. Results indicate that discretionary users exhibit expert-like characteristics on a small set of routine editing tasks, beyond which their behavior is quite novice-like. A procedural model that indicates areas where software design could improve the editing performance of discretionary users is also discussed.
Gestures with Speech for Graphic Manipulation BIBA 231-249
  Alexander G. Hauptmann; Paul McAvinney
This paper reports on the utility of gestures and speech to manipulate graphic objects. In the experiment described herein, three different populations of subjects were asked to communicate with a computer using either speech alone, gestures alone, or both. The task was the manipulation of a three-dimensional cube on the screen. They were asked to assume that the computer could see their hands, hear their voices, and understand their gestures and speech as well as a human could. A gesture classification scheme was developed to analyse the gestures of the subjects. A primary objective of the classification scheme was to determine whether common features would be found among the gestures of different users and classes of users. The collected data show a surprising degree of commonality among subjects in the use of gestures as well as speech. In addition to the uniformity of the observed manipulations, subjects expressed a preference for a combined gesture/speech interface. Furthermore, all subjects easily completed the simulated object manipulation tasks.
   The results of this research, and of future experiments of this type, can be applied to develop a gesture-based or gesture/speech-based system which enables computer users to manipulate graphic objects using easily learned and intuitive gestures to perform spatial tasks. Such tasks might include editing a three-dimensional rendering, controlling the operation of vehicles or operating virtual tools in three dimensions, or assembling an object from components. Knowledge about how people intuitively use gestures to communicate with computers provides the basis for future development of gesture-based input devices.
Analysis of Expert Reasoning in Hardware Diagnosis BIBA 251-280
  Nancy E. Reed; Paul E. Johnson
We present an analysis of expert reasoning in the domain of computer hardware diagnosis. The methods used in the study include directed interviews, observation and techniques of protocol analysis. The task investigated was the diagnosis of a complex piece of computer hardware. The initial symptoms are usually insufficient to determine the cause of a fault, requiring the acquisition of more data. Many thousands of pieces of data may be sought, but there is only enough time to obtain a few. We found that the experts use strategies to focus and obtain only the most relevant of this data. In addition, they use models of the hardware containing the diagnostically useful information. The strategies used are both powerful and efficient. Although they are specialized and save time and effort in performing the task, they are also general enough to apply to the majority of the problems encountered, including novel faults. As a test of these ideas, a prototype expert system was implemented. It is a model of problem solving and serves as a test of the adequacy of the knowledge described to perform diagnosis.
SOLA*: Students On-Line Advisor BIBA 281-312
  F. N. Arshad; G. Kelleher
This paper is about educational advising and adaptation to learner style in the provision of advice about learning. The recipients of the advice are students attempting to complete a course of study. SOLA*, an advisory system (implemented in SMALLTALK-80) provides help with what should be learned, how it should be learned and when it should be learned. Providing advice of this sort is becoming increasingly important as the range of educational materials available to the student continues to increase and the need for independent co-ordination of study becomes more common. Advisory systems are valuable in circumstances where learners are expected to control their own learning as not all students (particularly at first) are capable of successfully managing their study time. SOLA* is flexible in its approach to advising about study plans not only at the level of local decision making (e.g. which study materials to use) but also at the level of its global approach to a domain, SOLA* provides a series of study plans adapted to the learner.
A Lexical Analyser for Arabic BIBA 313-330
  Allel Feddag; Eric Foxley
Practical natural language processing (NLP) systems such as database front-ends, deductive databases or object-oriented databases are at the forefront of research into the next-generation intelligent database systems. The research described in this paper has been aimed at integrating front-end paradigms and rule-based deduction to provide a single powerful framework for database systems in Arabic. The lexicon stores only roots of verbs and uses a program intelligent enough to handle all derived forms automatically. This is significant, as these alone represent 70% of the total dictionary. As part of the discussion of this system, its utility in such NLP applications as parsing and machine translation is examined.
Characterizing Advisory Interactions between Statisticians and Entomologists BIBA 331-345
  Rye Senjen; Paddy Austin
An advisory expert system needs to model the expert's problem-solving behaviour as well as the interaction between advisor and client. An empirical study is reported that investigates advisory-client interactions in the field of insect sampling plan design. A form of discourse analysis was used to ascertain the roles of the participants and to develop a model of the functions performed during advice giving. The study used simulated interactions. Analysis of the interactions showed that the advisor controlled the relationship. The interaction can be characterized by a three-part model: information collecting, advice giving and closing. In three quarters of the interactions the advisor cycled between the information collecting and advice-giving episodes. This interchanging among episodes indicates a refinement of the advice as the statistician collects more information.

IJMMS 1993 Volume 38 Issue 3

Applying Gentner's Theory of Analogy to the Teaching of Computer Programming BIBA 347-368
  Yam San Chee
This research empirically tests the postulations of Gentner concerning the properties of explanatory analogy. It does so in the context of teaching programming. The factor analogy was operationalized by varying the clarity and systematicity/abstractness of the analogies used. The dependent variables were score obtained on program comprehension and program composition tasks and the time taken to perform the tasks. Research subjects were 15- to 17-year-olds without prior exposure to computer programming. Differences in age were controlled. The results provide empirical support for Gentner's postulations on the relative goodness of competing analogies. In particular, good explanatory analogies are characterized by clarity and high systematicity/abstractness.
An Experimental Investigation of Interface Design Alternatives: Icon vs. Text and Direct Manipulation vs. Menus BIBA 369-402
  Izak Benbasat; Peter Todd
This paper reports on two experiments which examine the effects of iconic and direct manipulation interfaces on the performance of casual users using an electronic mail system. There are two key aspects to these experiments. First, they have been carefully designed to separate the effect of iconic representation from that of direct manipulation in order to examine the independent effect of each as well as their joint effect. Second, subjects performed the same experimental task three different times over 1 week, thus allowing for the effects of icons and direct manipulation interfaces to be assessed over repeated trials. Each experiment measured time taken and errors made in task completion as dependent variables.
   Results indicate that there were no advantages associated with iconic representations compared to text-based representations of actions and objects. Subjects working with direct manipulation interfaces completed the task faster than those with menu-based interfaces. However, this difference in time was not significant when the task was repeated for a third time, indicating that the benefits to direct manipulation might diminish after a learning period. No interface was better than others in terms of reducing error rates when interacting with the computer system.
A Dynamic Reliability Technique for Error Assessment in Man-Machine Systems BIBA 403-428
  P. C. Cacciabue; A. Carpignano; C. Vivalda
This paper presents a methodology still under development for the analysis of human errors, named DREAMS (Dynamic Reliability technique for Error Assessment in Man-Machine Systems), which is dedicated to human reliability analysis and which identifies the origin of human errors in the dynamic interaction of the operator and the plant control system. This distinctive aspect differentiates DREAMS from the most commonly applied techniques in the nuclear as well as the conventional industries. Indeed, in the proposed methodology, the human behaviour depends on the working environment in which the operator acts ("external world"), i.e. the control room, and on the "internal world", i.e. his psychological conditions, which are related to stress, emotional factors, fixations, as well as to lack of intrinsic knowledge. As a logical consequence of the dynamic interaction of the human with the plant under control, either the error tendency or the ability to recover from a critical situation may be enhanced. The probability of erroneous actions and of recovery is thus a function of a generic "stress-in-action" correlation, which represents the effect of the internal-external worlds on the operator behaviour. This leads to the evaluation of Instantaneous Human Error/Recovery Probabilities (IHEP and IHRP) which are dynamically evaluated during the unfolding of the sequence of the man-machine interaction. These can then be used for evaluating an overall probability measure of plant safety related to human erroneous actions. A study case of application of the methodology to a control system of a real nuclear power plant is given in the paper as a sample case.
What Do You Expect to Get When You Ask for "A Cup of Coffee and a Muffin or a Croissant"? On the Interpretation of Sentences Containing Multiple Connectives BIBA 429-434
  Judith Avrahami; Yaakov Kareev
Sentences containing both the AND and the OR connectives (e.g. A and B or C) may be interpreted either as "(A and B) or C" or as "A and (B or C)". The present study explored the effects of two variables -- the order of the two connectives and category membership of the arguments connected -- on the interpretation of such sentences. Some 160 undergraduates had to choose between alternative interpretations of a sentence involving three arguments connected by AND and OR. It was found that in most cases AND is considered to bind the arguments more strongly than OR. However this tendency is order- and content-sensitive.
An Object-Oriented Language for Distributed Artificial Intelligence BIBA 435-453
  Giovanni Adorni; Agostino Poggi
In this work, we are presenting an object-oriented language to define distributed artificial intelligence (DAI) systems. This language allows for the description of control and communication management, a hierarchical representation of a DAI system at different levels of detail, and the simulation of its performance.
   The language is based on three different entities -- sequential actors, distributed actors and channel managers. A sequential actor describes a single DAI module (agent). A distributed actor represents a net of agents characterizing both a whole DAI system and its sub-parts. A channel manager administers channel communications among agents.
   This paper specifically illustrates (i) how data and procedural abstraction properties of such a language satisfy all the representation requirements for modelling DAI systems, and (ii) how message passing in an access-oriented paradigm can be an excellent basis for the representation of the mechanisms which drive the performance of a DAI system. We conclude with an example of the language application.
Focusing Based on the Structure of a Model in Model-Based Diagnosis BIBA 455-474
  P. Nooteboom; G. B. Leemeijer
The high computational complexity of existing methods for model-based diagnosis imposes a limit on the application of those methods in practical situations. Therefore, a lot of research time is spent on methods that improve the tractability of model-based diagnosis. In this paper, a new method for improving the tractability of model-based diagnosis is presented. The reasoning mechanism used is restricted to a subset of the components in the model called the focus. This focus is selected during diagnosis, based on (1) the structure of the model and the observed values at the system outputs, and (2) measurable connections in the model. Experiments have shown that focusing improves the practical applicability of model-based diagnosis. Furthermore, we will discuss a method for ordering the resulting set of diagnoses in a way analogous to determining a focus, i.e. based on the structure of the model and observations available.
User Acceptance of Information Technology: System Characteristics, User Perceptions and Behavioral Impacts BIBA 475-487
  Fred D. Davis
Lack of user acceptance has long been an impediment to the success of new information systems. The present research addresses why users accept or reject information systems and how user acceptance is affected by system design features. The technology acceptance model (TAM) specifies the causal relationships between system design features, perceived usefulness, perceived ease of use, attitude toward using, and actual usage behavior. Attitude theory from psychology provides the rationale for hypothesized model relationships, and validated measures were used to operationalize model variables. A field study of 112 users regarding two end-user systems was conducted to test the hypothesized model. TAM fully mediated the effects of system characteristics on usage behavior, accounting for 36% of the variance in usage. Perhaps the most striking finding was that perceived usefulness was 50% more influential than ease of use in determining usage, underscoring the importance of incorporating the appropriate functional capabilities in new systems. Overall, TAM provides an informative representation of the mechanisms by which design choices influence user acceptance, and should therefore be helpful in applied contexts for forecasting and evaluating user acceptance of information technology. Implications for future research and practice are discussed.
The Effects of Alternative Forms of Knowledge Representation on Decision-Making Consensus BIBA 489-507
  Jack Arthur, Jr. Gowan; Charles W. McNichols
This research compared three different forms of feedback in a Delphi exercise, and demonstrated the consensus building effectiveness of feedback represented by if-then rules. The rules, similar to those employed in expert system knowledge bases, were induced from experts' decisions on a sequence of test cases and used to summarize the group's decision policy. Consensus was measured in terms of pairwise subject agreement after each round of decisions, and the measurement was obtained from panels of experts making decisions with three forms of feedback: if-then rules, frequency distributions or regression weights. The going concern decision scenario provided test cases derived from small business data and required evaluation of the likelihood of a company's economic viability 1 year in the future. Subjects were experienced bank loan officers and small business development center consultants. A three-round Delphi experiment accomplished by mail showed that if-then rules were more effective in motivating consensus than either frequency distributions or regression weights. As a result of this research, a prototype Group Decision Support System (GDSS) has been implemented to facilitate repetitive group decisions such as those required in evaluating loan applications or capital investment projects.
Operationalizing the Opportunistic Behavior in Model Formulation BIBA 509-540
  Ajay S. Vinze; Arun Sen; Shue Feng T. Liou
A review of the literature in model formulation reveals varying approaches. These approaches can be classified under efforts related to design and planning. A quasi-experiment was conducted using the think-aloud approach to study the process of model formulation as undertaken by expert modelers. Model in the present context refers to a structure which is built purposely to exhibit features and characteristics of certain scenarios. Model formulation then is the human problem solving activity focused on the process of constructing models. The domain of interest was managerial problem solving, in particular, problems from the areas of Auditing, Forecasting and Production Planning were studied. An evaluation of the verbalizations by the experts reveals the predominant use of opportunism in the process. Opportunism is one approach to problem solving where the modeler follows the most promising leads at any point in time. A cognitive model was used to categorize and understand the expert protocols as derived from the verbalizations. A comparison of the verbalizations using this model shows similarity in the process of formulation. The cognitive model is operationalized using a blackboard architecture.

European Association for Cognitive Ergonomics: Book Reviews

"The Psychology of Menu Selection: Designing Cognitive Control at the Human/Computer Interface," by Kent L. Norman BIB 541-546
  Rarrell R. Raymond; Eric S. Lee
"Languages for Developing User Interfaces," edited by B. A. Myers BIB 541-546
  N. P. Rousseau
"A Theory of Computer Semiotics: Semiotic Approaches to Construction and Assessment of Computer Systems," by P. Bogh Andersen BIB 541-546
  Heather Stark
"Effective Colour Displays: Theory and Practice," by D. Travis BIB 541-546
  James Tresilian

IJMMS 1993 Volume 38 Issue 4

User Representations of Computer Systems in Human-Computer Speech Interaction BIBA 547-566
  Rene Amalberti; Noelle Carbonell; Pierre Falzon
We report an experiment designed to study whether models of human-human voice dialogues can be applied successfully to human-computer communication using natural spoken language. Two groups of six subjects were asked to obtain information about air travel via dialogue with a remote "travel agent". Subjects in the computer group were led to believe they were talking to a computer whereas subjects in the operator group were told they were talking to a human. Both groups of subjects actually talked to the same human experimenter. The study focuses on subjects' representations of interlocutor skill and knowledge, and differs from previous analogous studies in several respects: the task is more complex, giving rise to structured exchanges in natural language rather than to question/answer pairs in simplified language; specific attention has been paid to the design, which attempts to avoid biases that have flawed other studies (in particular, conditions are identical for both groups); the time factor has been taken into account (subjects take part in three sessions, at 1-week intervals).
   Some results confirm those of the literature, namely that subjects of the computer group tend to control and simplify their use of language more than those in the operator group. However, most observations are either new or in contradiction with previous results: subjects in the computer group produce more utterances but no significant differences were observed with respect to most structural and pragmatic features of language; the time factor plays a dual role. Subjects in both groups tend to become more concise. Operator group strategies differ significantly across sessions as regards scenario processing (problem solving) whereas computer group strategies remain stable.
   These differences in behavior between groups are ascribed to differences in representations of interlocutor ability.
User Modelling for Graphical Design in Complex Dynamic Environments: Concepts and Prototype Implementations BIBA 567-586
  Gunilla A. Sundstrom
In this paper, it is proposed that using knowledge-based technology to generate different types of models of a target domain can better assist designers of support systems. The main argument is that designers' decisions should be based on knowledge about a domain rather than on their "common sense" judgments. This approach is illustrated by describing a canonical user model capturing operators' information and knowledge acquisition behaviour. In the present work, this model is used to help designers identify what process variables need to be associated with flow diagrams representing the structure of the technical system. Concepts underlying the user modelling approach as well as the knowledge elicitation and knowledge representation process are described. Finally, functionality provided by the user model is discussed as well as how the approach can be evaluated.
Mental Models: Concepts for Human-Computer Interaction Research BIBA 587-605
  Nancy Staggers; A. F. Norcio
In interacting with the world, people form internal representations or mental models of themselves and the objects with which they interact (Norman, 1983a). According to Norman, mental models provide predictive and explanatory powers for understanding the interaction. More abstractly, Gentner and Stevens (1983) propose that mental models focus on the way people understand a specific knowledge domain. More concretely, Carroll (1984) views mental models as information that is input into cognitive structures and processes. What are mental models? Are they always formed? When formed, what are their characteristics? What are the functional consequences of having no model (if that is possible), an immature model, or a mature model? This paper intends to explore these questions.
AERIAL: Ad-Hoc Entity-Relationship Investigation and Learning BIBA 607-623
  L. M. Burns; A. Malhotra; G. Sockut; K.-Y. Whang
This paper discusses Browsing, an alternate style of working with databases, that facilitates the unplanned exploration of the structure and contents of a database by a novice user. We argue that Browsing promotes learning and helps bring the user's mental model of the problem space into correspondence with the model stored in the database. Browsing is illustrated through a learning scenario using AERIAL, a facility that allows the user to browse an Entity-Relationship database. The facilities provided by AERIAL are discussed in detail as well as their effectiveness in building the above correspondence.
The Use of Analytical Models in Human-Computer-Interface Design BIBA 625-660
  Leo Gugerty
Researchers in the human-computer interaction field have advocated that interface designers use analytical models of the user (e.g. the GOMS model) to help them consider user needs during the design process. This paper surveys the literature on analytical models for interface designers, focusing initially on empirical studies of the validity of these models. This survey shows that analytical models can by used by interface designers in two ways: (1) as task analytic tools that help in generating preliminary design ideas, and (2) as tools for evaluating preliminary designs by predicting user performance and satisfaction. Empirical studies have demonstrated that analytical models can be used in the task analysis phase of design; in these studies, models were successfully used to generate new designs for existing interfaces, with the new designs leading to improved user performance.
   Regarding the use of analytical models to evaluate interface designs, a number of empirical studies have shown that these models can help explain the factors affecting user performance in a precise, quantitative fashion. However, these explanations have only been given for existing interfaces. Advocates of analytical models have not yet demonstrated convincingly that models can generate accurate a priori predictions of user performance for a new interface. This kind of prediction is necessary before analytical models can be used effectively by interface designers.
   The final section of the paper focuses on the practical constraints affecting the use of analytical models in interface design organizations, such as organizational schedules, budgets and training requirements. In this section, suggestions are made concerning research needed before analytical models can be used in real-world design projects.
A New Approach to Detecting Missing Knowledge in Expert System Rule Bases BIBA 661-688
  Alun D. Preece
Two of the most important and difficult tasks in building expert systems are knowledge acquisition (KA) and quality assurance (QA). QA involves verification and validation (V&V) techniques, which often reveal errors that must be rectified through further KA. Traditionally, V&V is done by extensive testing of the expert system, but this is difficult, time-consuming, labour-intensive (requiring considerable human expert involvement) and unreliable (due to the large size of most expert system input domains). V&V can be assisted by using a verification tool to detect anomalies in the knowledge base of an expert system, such as redundancies, conflicts, circularities and deficiencies (missing knowledge). Such anomalies are usually indicative of errors in the expert system. Deficiencies are important for QA and KA: they indicate incomplete portions of the knowledge base which should either be specified as known domain bounds or completed via further KA. This paper describes the COVER deficiency detection tool which: minimizes human expert involvement in knowledge base verification; focuses the search for meaningful deficiencies; integrates closely with checks for redundancy, conflicts and circularity; maximizes user-control over deficiency detection; and overcomes the combinatorial explosion traditionally associated with the deficiency check. The paper describes how COVER uses heuristics about the nature of likely deficiencies to improve its performance and clarify reporting of deficiencies to the user. COVER performance is analysed in detail, both theoretically and on real-world expert system knowledge bases.
Evaluating User-Computer Interaction: A Framework BIBA 689-711
  M. Sweeney; M. Maguire; B. Shackel
A framework is described, which classifies usability evaluations in terms of three dimensions; the approach to evaluation, the type of evaluation and the time of evaluation in the context of the product life cycle. The approaches described are user-based, theory-based and expert-based. The approach to evaluation reflects the source of the data which forms the basis of the evaluation. The types of evaluation are diagnostic, summative and metrication. These reflect the purpose of the evaluation and therefore the nature of the data and likely use of the results. The time of testing reflects the temporal location in the product life cycle at which the evaluation is conducted. This dictates the representation of the product which is available for evaluation.
   The paper describes the relationship between these three framework dimensions. It also relates the methods of data capture, measurements and criteria which may be appropriately applied in various evaluation contexts. The latter part of the paper focuses on a more detailed review of methods which are associated with the most commonly applied and often most effective approach, i.e. the user-centred diagnostic evaluation.
   Finally the paper considers the need to perform evaluations more effectively in the design of products and systems in the commercial world. The discussion addresses the need for computer support tools to facilitate the handling of resulting data from user trials.
Performance Effect of Matching Computer Interface Characteristics and User Skill Level BIBA 713-724
  James E. Trumbly; Kirk P. Arnett; Merle P. Martin
An experiment was conducted to see if matching user interface characteristics to user computer knowledge could improve user performance. User performance was measured by simulation game results, the extent of errors and user response time. Experiment subjects were tested to determine their level of computer knowledge using a pretested instrument. The subjects were randomly assigned to either a novice or experienced oriented interface version of a space manufacturing simulation. Results show that task performance (manufacturing profit) significantly increased when user interface characteristics were matched to user computer knowledge (e.g. novice interface to novice user). The relationship between interface match and user errors was marginally significant. The results of this study are important to software designers and practitioners purchasing software products.
Artificial Intelligence in Education: Using State Space Search and Heuristics in Mathematics Instruction BIBA 725-746
  Anthony E. Kelly; D. H. Sleeman; Kenneth J. Gilhooly
Two powerful techniques from AI -- the use of heuristics, and state space analysis -- are currently influencing the practice of pedagogy, particularly in mathematics. The current study compared the performance of college students on a mathematics task as a function of their exposure to each of these techniques. Specifically, it tested the hypotheses that instruction in the use of AI-type cued heuristics, and state space analysis (search) would improve students' solving of trigonometric identities (the target task), and their solving of group-theory identities (the transfer task). Four conditions were developed in which the effects of cued heuristics alone, search alone, search and cued heuristics, and neither search nor cued heuristics were tested. A significant main effect of cued heuristics was found, but the positive effects of using cued heuristics did not transfer to problems involving group theory identities. Little gain was found for those using state space analysis. Implications for the use of these techniques in mathematics instruction are discussed.

IJMMS 1993 Volume 38 Issue 5

Experimental Comparison of Navigation in a Galois Lattice with Conventional Information Retrieval Methods BIBA 747-767
  Robert Godin; Rokia Missaoui; Alain April
A controlled experiment was conducted comparing information retrieval using a Galois lattice structure with two more conventional retrieval methods: navigating in a manually built hierarchical classification and Boolean querying with index terms. No significant performance difference was found between Boolean querying and the Galois lattice retrieval method for subject searching with the three measures used for the experiment: user searching time, recall and precision. However, hierarchical classification retrieval did show significantly lower recall compared to the other two methods. This experiment suggests that retrieval using a Galois lattice structure may be an attractive alternative since it combines a good performance for subject searching along with browsing potential.
Preferred Mental Models for Direct Manipulation and Command-Based Interfaces BIBA 769-785
  Ray E. Eberts; Kalappa P. Bittianda
One of the main methods used to compare direct manipulation and command-based interfaces is to examine user preferences. User preferences are extended in our research to examine which mental model, direct manipulation or command-based, subjects prefer to use in transfer situations. One group of subjects was provided with both a command-line and direct manipulation interface during a training phase. After training, several transfer experiments were conducted to determine the preferred mental model. The developed mental models were investigated by "running" the models and extending the models to new situations. They were evaluated by determining the operators specified by the subjects. By comparing the class of operators to subjects trained on only one of the interfaces, direct manipulation or command-based, the preferred model could be determined. The preferred mental model for "running" the model was the direct manipulation. For extensions that had a concrete or graphical basis in the interface, the direct manipulation was preferred. For extensions that were abstract, one model was not preferred over the other. Some ability to use multiple mental models, based upon the task, was also observed.
Q-Analysis of User-Database Interaction BIBA 787-803
  Thomas L. Jacobson; David S. Fusani; Wenjie Yan
This study jointly employed two research approaches in developing a process oriented methodology for studying information seeking during user-database interaction. A "time-line" structured interview technique was used to gather 26 novice-user accounts of search sessions on NEXIS. These accounts included a description of the search session, questions that occurred to users during this session, and indications on whether answers were obtained. A 17 category descriptive content analysis scheme was developed and Q-analysis was performed on the accounts as coded into this scheme. Results provide a rich description of the search process from the user's viewpoint. Searchers below the mean on a document relevance measure most frequently had questions after being in some way rebuffed by the system. Above mean searchers were also active questionners, but their questions were most frequently associated with string design rather than being rebuffed. The above mean group also rated their questions as less important than did the below mean group. In addition, above mean searchers were more active users of system help features and information provided onscreen in general.
A Uniform Graphical View of the Program Construction Process: GRIPSE BIBA 805-837
  K. Halewood; M. R. Woodward
This paper describes the ideas leading to the development and construction of a Graphical Integrated Programming Support Environment (GRIPSE) for a modified form of the Pascal language. In order to represent all procedural and declarative aspects of a Pascal program, the original Nassi-Shneiderman diagrams forming the graphical basis of GRIPSE have been augmented into a multi-level, three dimensional system capable of supporting all static and most dynamic aspects of a program in a uniform manner. The dynamic view exists by virtue of an interpreter that has been incorporated along with debugging aids. A planned extension to the environment is an interactive form of mutation testing called Firm mutation which exploits the middle-ground between Strong and Weak mutation, both of which have traditionally been applied in a non-conversational, non-interactive mode of use.
The Effects of User Involvement: Some Personality Determinants BIBA 839-855
  Stephen R. Hawk
A number of studies have been conducted with the purpose of confirming or disconfirming the general argument that user involvement in information system development is beneficial. Few studies have been undertaken, however, to support the view that the benefits of user involvement are affected by conditions surrounding systems development. This paper presents the results of a field study investigating the moderating influence of four user personality characteristics on the effects of user involvement on user information satisfaction. The relationship between user involvement and user information satisfaction was the most positive for users who were tolerant of ambiguity, and for users with a high need for achievement. An unexpected finding was that the relationship between user involvement and user information satisfaction was more positive for users with external-control beliefs than for users with internal-control beliefs. Degree of users' dogmatism had no impact on the outcome of user involvement.
A Probabilistic Inference Scheme for Hierarchical Buggy Models BIBA 857-872
  N. Tokuda; A. Fukuda
The probabilistic reasoning scheme of Dempster-Shafer theory provides a remarkably efficient bug identification algorithm for a hierarchical Buggy model. In the particular Buggy model generated by the repair theory of Brown & Van Lehn (1980, A generative theory of bugs in procedural skills, Cognitive Science, 2, 155-192), both Shafer & Logan (1987, Implementing Dempster's rule for hierarchical evidence, Artificial Intelligence, 33, 271-298) and Gordon & Shortliffe (1985, A method for managing evidential reasoning in a hierarchical hypothesis space, Artificial Intelligence, 26, 324-357) schemes have provided almost identical computational accuracy although the latter involves an approximation of a "smallest superset". If n denotes the number of bugs to be identified, the computational complexity of the two schemes, originally of O(n4/3) and O(n²) respectively, can be improved to O(n) using the simplified top-down calculation scheme whereby from among all the nodes we first locate the particular "parental" node to which the bug belongs and then the bug itself among the sibs within the node. On average, about 5-7 problems are adequate to raise the belief function of the bug to 95% level, based on the evidential reasoning schemes.
User Interface Design for Cooperating Agents in Industrial Process Supervision and Control Applications BIBA 873-890
  Nicholas M. Avouris; Marc H. Van Liedekerke; Georgios P. Lekkas; Lynne E. Hall
Computer systems based on cooperating agent architectures are currently introduced in industrial process supervision and control applications as operator support systems in tasks such as fault diagnosis, system restoration etc. Cooperating agents are relevant to these applications since they involve a high degree of physical distribution, the operators' decisions are often based on multiple conflicting views which can be moderated by the cooperating agents, and the domains are complex with high degree of modularity.
   In the frame of the research project ARCHON, a multi-agent system architecture has been defined in order to be used in industrial process supervision and control applications. An issue of high importance in this context is the interaction with the human operator and the design of a user interface which supports this interaction. The impact of multi-agent architectures on the user interface design of industrial process supervision and control systems is the subject of this paper. An example of an interface designed for a multi-agent system in the area of electrical network supervision is provided.

IJMMS 1993 Volume 38 Issue 6

NALIGE: A User Interface Management System for the Development of Natural Language Interfaces BIBA 891-921
  Bill Z. Manaris; Wayne D. Dominick
One of the major problems encountered by the average computer user is the syntactic inflexibility and linguistic opacity of existing operating system command languages. These characteristics, which are manifested in terms of rigid syntactical rules and cryptic/obtuse acronyms, have been shown to result in significant loss of productivity among novice and intermediate users. This paper investigates one approach to addressing this problem, namely the introduction of natural language interfaces placed on top of the standard operating systems command language interpreters. Specifically, it presents a user interface management system which accepts a set of well-formed specifications and produces an autonomous natural language interface exhibiting the prescribed linguistic behavior. These specifications describe the linguistic knowledge to be encapsulated in the target interface in terms of the lexical, syntactic, semantic, and pragmatic elements of the application domain. Additionally, this paper discusses a systematic methodology to be used in identifying the linguistic elements of a given application domain. This methodology consists of a set of formal techniques, computational tools, and three knowledge-base specification languages to be employed by the interface designer during all stages of development and subsequent maintenance of a specific natural language interface.
Human Local Representations of Uncertainty: A Methodology for and Results from Comparing Different Schemes for Representing Uncertainty BIBA 923-947
  Bruce M. Perrin; David S. Vaughan; Robert M. Yadrick; Peter D. Holden
Human reasoning under uncertainty has been shown to consist of a series of "local computations" in which a complex problem is broken into a series of simpler decisions. For practical and epistemological reasons, expert systems which must reason under uncertainty have taken a similar approach. Thus, choice of an appropriate Uncertainty Representation Scheme (URS) is dependent upon the robustness of the scheme to variations in the local computations that are expected in the application area. Researchers have based their claims for robustness of a particular URS on the apparent match between the scheme's stated assumptions and the expected "statistical" characteristics of the local computations. The limited empirical data that are available suggest that these claims may be ill-founded. Rather, this paper argues that a scheme's robustness should be measured relative to humans' local computations which are used to solve the problem, and we present a methodology for doing so. Application of this methodology suggests that representations of local computations are strongly influenced by the choice of a URS and that the accuracy of the resulting solutions are substantially influenced. Implications for full scale applications and directions for further research are also discussed.
An Approach to Intelligent Automated Window Management BIBA 949-983
  Douglas J. Funke; Jeannette G. Neal; Rajendra D. Paul
The CUBRICON Intelligent Window Manager (CIWM) is a knowledge-based system that automates windowing operations. The CIWM is a component of CUBRICON, a prototype knowledge-based multi-media human-computer interface. CUBRICON accepts inputs and generates outputs using integrated multiple media/modalities including speech, printed/typed natural language, tables, forms, maps, graphics, and pointing gestures. The CIWM automatically performs window management functions on CUBRICON's color and monochrome screens. These functions include window creation, sizing, placement, removal, and organization. These operations are accomplished by the CIWM without direct human inputs, although the system provides for user override of the CIWM decisions.
   The motivation for automated window management is based on the premise that, by freeing the user's cognitive and temporal resources from the task of managing the human-computer interface, more of these resources are available for the user's application domain activities. As the problems and tasks confronting computer users become more complex and information intensive, the potential of this approach for improving overall performance is enhanced. Recent research discussed in this paper indicates that, for some database management tasks, a significant portion of the user's time is spent in managing the window-based interface. If these findings are representative of the larger range of computer-based tasks that use windowing systems, the concept of automated window management offers great potential for enhancing human performance on these computer-based tasks.
   This paper provides a brief overview of the CUBRICON system and describes the CIWM and its underlying design principles and premises. The following important CIWM features are discussed: the hybrid tiled and overlapped approach to window layout; an algorithm for determining the importance of a window based on its contents, relation to the ongoing dialogue, time of creation, frequency of use, and recency of use; and an approach to determining window size based on clutter and object resolution requirements. Actual interactive examples are provided to illustrate the CIWM functionality. Results of an evaluation of CUBRICON support the design. Those results which pertain specifically to the CIWM are presented. Limitations and applicability of this research are also discussed.
Why Can't Smart Students Solve Simple Programming Problems? BIBA 985-997
  Russell L. Shackelford; Albert N. Badre
Computer programming education is evolving on several fronts. This study investigated aspects of the question: how adequate are our conceptions of how to best teach the subject matter? Previous research suggested that the teaching of programming should be focused on "problem solving strategies" ("Model A") rather than on the syntactic/semantic aspects of writing programs ("Model B"). This study was designed to test for student programming performance differences based upon feedback content obtained from each of these two models. The results indicate that certain Pascal loop construct definitions are too general with respect to loop construction. A stricter loop schema resulted in superior performance. The findings also argue that the WHILE construct should receive stricter instructional treatment. In addition, a "constructive" approach to "Model A" feedback (focusing on programmer processes) correlated with subsequent performance superior to that of the control group, whereas neither the conventional approach to "Model A" feedback nor "Model B" feedback (focusing on programmer errors) did so. This finding argues that an approach which relies on narrative treatments and/or error messages is ineffective and that constructive decision rules should serve as a basis for feedback generation and perhaps other aspects of teaching.
The Social and the Cognitive in Human-Computer Interaction BIBA 999-1016
  R. J. Anderson; C. C. Heath; P. Luff; T. P. Moran
Of late, designers of interactive systems and other exponents of HCI have expressed an increased interest in the contribution which Social Science might make to design. Using recent discussions of "distributed cognition" as our stalking horse, we show that a strategy of simple annexation or incorporation is unlikely to realize the value which the Social Sciences might contribute. Such value will not be derived by a "filling out" of design requirements through the addition of social dimensions to cognitive ones. Rather, it will take the form of a re-appraisal of deep-seated distinctions such as that between the social and the cognitive. In the context of some on-going work at EuroPARC, we examine the possibilities which this re-appraisal might offer. We conclude with a review of the implications of this kind of re-appraisal for the design of interactive systems.
Searching Information in Hypertext Systems Using Multiple Sources of Evidence BIBA 1017-1030
  Jacques Savoy
In proposing a searching strategy in a hypertext environment, we have considered three criteria: (1) the retrieval process should use multiple sources of evidence, including multiple indexing schemes and multiple search mechanisms; (2) the hypertext links should be exploited in order to find more relevant material and "good" starting points for browsing; and (3) the user's information need must be easily expressed. To satisfy these three criteria, we have implemented a hypertext system where the user, instead of writing a formal query, just selects relevant terms and/or documents. Based on multiple indexing schemes, Bayesian networks are used to manage the indexing spaces and to store the user's information need. To combine multiple search techniques, the hypergraph is also composed of implicit links (bibliographic coupling, co-citation, etc.), and computed links storing the nearest neighbors of each node. Using link semantics, a constrained spreading activation retrieves relevant nodes for browsing.
Supporting Pascal Programming with an On-Line Template Library and Case Studies BIBA 1031-1048
  Patricia K. Schank; Marcia C. Linn; Michael J. Clancy
We propose a template library as a good representation of programming knowledge, and programming case studies as part of an effective context for illustrating design skills and strategies for utilizing this knowledge. In this project, we devised an on-line network of Pascal programming templates called a template library, and tested it with subjects (classified as novice, intermediate, and expert Pascal programmers) both as a stand alone resource and in conjunction with programming case studies. We investigated three questions using these tools: 1) How do subjects organize templates? 2) How well can subjects understand and locate templates in the template library? 3) Does the template library help subjects reuse templates to solve new problems? Results suggest that the template representations helped subjects remember and reuse information, and that subjects gained deeper understandings if the representation was introduced in the context of a programming case study.

European Association for Cognitive Ergonomics: Book Reviews

"User Interface Design," by H. Thimbleby BIB 1049-1055
  Thomas Green
"MARCEL: Simulating the Novice Programmer," by James C. Spohrer BIB 1049-1055
  Hank Kahney
"Eliciting and Analyzing Expert Judgement, A Practical Guide," by M. Meyer and J. Booker BIB 1049-1055
  N. Shadbolt
"Information Modelling Practical Guidance," by Richard Veryard BIB 1049-1055
  David Benyon