P. Szekely, P. Luo, and Neches, R. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In CHI '92: Human Factors in Computing Systems, Monterey, CA, May 1992, pp. 507--515.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....specifications into models from which a running UI can be derived. It is useful to be able to specify a desired UI formally into models before building it, particularly if a mockup suitable for testing can be obtained directly from the specification [1,9,10] Each such model has three properties [16,18]: 1. Declarative: specifications contained in each model should be expressed in a declarative specification language called interface modeling language. 2. Editable: specifications can be edited either manually by a human operator or automatically by an automaton. 3. Analyzable: specifications ....

Szekely, P., Luo, P., Neches, R.: Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In: Proc. of ACM Conference on Human Factors in Computing Systems CHI'92 (Monterey, May 3-7, 1992). New York, ACM Press (1992) 507--515

....a high degree, in order to have executable interfaces in early design stages. During the further development the model specifications will become more elaborated, and therefore the need for automation will decrease, while the UI quality increases (see [32] With this approach inspired by HUMANOID [33], rapid prototyping is possible. Domain Model Elements Implementation e.g. Java classes, db tables, Domain Instance Elements Instance e.g. Classifier , Attribute , Use CaseModel Elements Implementation Instance Elements e.g. Java objects, db rows XML document Instance Elements ....

P. Szekely, P. Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The Humanoid Model of Interface Design. In CHI, pages 507--515, May 1992.

....presented here, and nor are specific notations or tools. MB UIDE References Local ADEPT [27, 23, 52] Queen Mary and Westfield College, UK AME [28] Fachhochschule Augsburg, Germany FUSE [38, 39, 25] Technische Universitat Munchen, Germany GENIUS [21] University of Stuttgart, Germany HUMANOID [44, 46, 26] University of Southern California, USA JANUS [2, 3] Ruhr Universitat Bochum, Germany ITS [50, 51] IBM T.J. Watson Research Center, USA MASTERMIND [47, 9, 43] University Southern California, Georgia Inst. Tech. USA MECANO [32] Stanford University, USA MODI D [35, 33, 34] Stanford University, ....

P. Szekely, P. Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In Proceedings of SIGCHI'92, pages 507--515, May 1992.

....specified as to how the defined model or models are to be translated into the graphical user interface. The designer may be allowed to specify one or more models, such as a domain model, a data model, a dialogue model, a platform model and a task model. Model approaches include UIDE [3] HUMANOID [16], ADEPT [12] TRIDENT [19] Fran [6] and Teallach [8] Most early automatic generation approaches are based on a single model and imposed a fixed set of interactors with a rigid style guide. Application specific interactors are not able to be specified. Recent research defines additional models, ....

P. Szekely, P. Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design Understanding and Supporting the Design Process. In Proceedings of ACM CHI'92 Conference on Human Factors in Computing Systems, 1992, pages 507-515.

....automatically the generation of interfaces from such specifications. Mecano [2] which is the precursor of MOBI D, allowed developers to generate form based interfaces from explicit models of an application s domain. ADEPT [1] drives interface generation from models of user s tasks. HUMANOID [4] emphasized presentation models as a basis for producing interfaces. Other approaches utilized Entity Relationship data models or Petri Nets as dialog models to produce interfaces. These approaches are limited in two aspects. First, by defining only part of the design knowledge of an interface, ....

Szekely, P., Luo, P., Neches, R, Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design, in [CHI92], pp. 507-514.

....KNOWLEDGE There has been considerable progress in the field of modelbased interface development in recent years [2] A number of systems have been implemented and have been successful at showing the potential of the model based paradigm. Some of the well known systems include ADEPT [5] HUMANOID [4], and Mecano [3] However, none of the available model based systems has implemented an effective process for eliciting user task models from domain experts, a key requirement for userinterface design. Some systems, including HUMANOID and Mecano, do not even have representations for user tasks in ....

Szekely, P., Luo, P., Neches, R. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design, in Proceedings of CHI'92, pp. 507-514.

....critics, simulators and model checkers. MB IDEs may be compared with respect to the number and expressiveness of their models, and the design tools which exploit the model based information. For comparisons of collections of proposals, see [15, 8] Early MB IDEs, such as UIDE [6] and Humanoid [17], incorporate relatively simple application, dialogue and presentation (i.e. user interface component) models, plus a rule based user interface generator. ADEPT [9] adds a model of user system tasks, at a higher level of abstraction than dialogue, to produce prototype interfaces for evaluation ....

....String dm PrimitiveType specifies tyro I dm Parameter definedin defines dm PrimitiveType dm Operation siamre ttions Figure 2. Domain 3.2.1 Basic Task Model Structure. The task model shares its basic structure with the task models of several MB IDEs such as Adept [10] Mastermind [17] and TADEUS [7] An example of a task model in the context of the Teallach tool is given in figure 3. The task model is a goal oriented task hierarchy, with its leaf nodes (termed primitive tasks) representing interaction or action tasks. The temporal relationship between sibling tasks is ....

P. Szekely, P. Luo, and R. Neches, "Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design", Proc. CH1 92, 1992, pp. 507-515.

....to specific controllers. For visual based interaction, BEACH uses an adapted version of the model view controller concept (see section 3. 5) Similarly, the document model separates the domain model (sometimes also called business domain objects or just model objects ) and the application model [15, 41]. Domain models represent entities of the domain world. Application models are used to describe all application aspects such as presentation and manipulation of domain objects. For, e.g. a text object, the domain model includes the string describing the text and text attributes like font or ....

Szekely, P., Luo, P., Neches, R. (1992). Facilitating the exploration of interface design alternatives the HUMANOID model of interface design. In Proceedings of the ACM Conference on Human factors in computing systems (CHI'92), pp. 507--515.

....specifications into models from which a running UI can be derived. It is useful to be able to specify a desired UI formally into models before building it, particularly if a mockup suitable for testing can be obtained directly from the specification [1,9,10] Each such model has three properties [16,18]: 1. Declarative: specifications contained in each model should be expressed in a declarative specification language called interface modeling language. 2. Editable: specifications can be edited either manually by a human operator or automatically by an automaton. 3. Analyzable: specifications ....

Szekely, P., Luo, P., Neches, R.: Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In: Proc. of ACM Conference on Human Factors in Computing Systems CHI'92 (Monterey, May 3-7, 1992). New York, ACM Press (1992) 507--515

....Design The discipline of model based user interface design advocates the explicit denotation of the conceptual abstractions that underlie a user interface. This formal description can then be used to drive the user interface at run time, and it can also serve additional purposes. Humanoid [Szek92, Szek93] and UIDE [Fole89, Suka93] are typical model based systems. I.1.1 Strengths The primary strength of user interface models is in their descriptive power and in their high level of abstraction. This is because they are geared towards the specification of user interface behavior, and need not cover ....

....record can be declared as a field name of type String, an address of type Address and a manager of type Employee. The interface genera 11 tor can then create a dialog box for display of such a field. The user interface generation process can be controlled by modifying the style rules. HUMANOID s [Szek92,Szek93,Luo93] model consists of commands, objects, global variables and data flow constraints. Commands have associated inputs (parameters) and preconditions for their applicability. An input describes one parameter of a command by defining the type, a predicate for semantic input validation and other ....

Pedro Szekely, Ping Luo, and Robert Neches. Facilitating the exploration of interface design alternatives: The HUMANOID model of interface design. In Proceedings of the ACM Conference on Human Factors in Computing Systems, pages 507--515, (Monterey, California, May 3-7) 1992.

....from declarative models: 1. Interpreters: the declarative models are interpreted directly, as in ITS [12] 2. UIMS Generators: the declarative models are mapped into alternative, generally lower level, representations of the interface, which are subsequently interpreted, as in Humanoid [11], TADEUS [4] and FUSE [8] 3. Source Code Generators: the declarative models are mapped to programs in an imperative programming language, as in JANUS [1] and later versions of Mastermind [10] which generate C code. Arguments can be made for and against each of the above approaches. For ....

Szekely, P., Luo, P., Neches, R. Facilitating the exploration of interface design alternatives: The HUMANOID model of interface design. In Proceedings of SIGCHI'92 (May 1992). 507515.

....presented here, and nor are speci c notations or tools. MB UIDE References Local ADEPT [27, 23, 52] Queen Mary and West eld College, UK AME [28] Fachhochschule Augsburg, Germany FUSE [38, 39, 25] Technische Universit at M unchen, Germany GENIUS [21] University of Stuttgart, Germany HUMANOID [44, 46, 26] University of Southern California, USA JANUS [2, 3] Ruhr Universit at Bochum, Germany ITS [50, 51] IBM T.J. Watson Research Center, USA MASTERMIND [47, 9, 43] University Southern California, Georgia Inst. Technology, USA MECANO [32] Stanford University, USA MODI D [35, 33, 34] Stanford ....

P. Szekely, P. Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In Proceedings of SIGCHI'92, pages 507-515, May 1992.

....of these two types of metrics will help prevent efficient, but ugly, interfaces as well as aesthetic, but inefficient interfaces. 2 Existing systems Several systems have incorporated metrics in various ways for evaluating user interfaces. Although others exist, such as GENIUS [3] HUMANOID [14], KRI AG [5] TRIDENT [16] these systems have less in common with the current research. Tullis conducted experiments to evaluate the effectiveness of a set of six task independent metrics for evaluating alphanumeric interfaces. Tullis metrics measure: the percentage of the display used to ....

Szekely, P., Luo, P, and Neches, R., Facilitating the exploration of interface design alternatives: The HUMANOID model of interface design, Proceedings of CHI'92 (1992), New York: ACM, 507-515.

....attributes. A second approach is to annotate the models used in automated program synthesis systems. The knowledge bases of specialized knowledge based synthesis systems, such as user interface development systems, can be augmented to support the generation of documentation and help as well [12]. Integration of I Doc with one or more such systems is an option being considered for future development. The third source of design information for documentation is code analysis. Analysis routines can detect components that appear to be creating objects, inserting into or removing from data ....

P. Szekely, P. Luo, and R. Neches. Facilitating the exploration of interface design alternatives: The HUMANOID model of interface design. In Proceedings of CHI'92, The National Conference on Computer-Human Interaction, pages 507--515, May 1992.

....represented in code but they are not normally accessible from outside the application code. An application model is accessible by the interface, the application and external tools at design and at run time. Examples of modelbased user interface management systems (UIMS) are UIDE [1,4] and HUMANOID [3,5] which both use a highlevel object oriented application model. ITS [6] is an example of a UIMS which encodes detailed interface style rules in order to generate very high quality user interfaces from the model. Model based UIMSs impose a design methodology on their users which requires them to ....

Szekely, P., P. Luo and R. Neches, Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design, Proceedings of CHI'92, ACM Conference on Human Factors in Computing Systems, Monterey, CA, pp. 507-515.

....since the language used in UIDE specifies the application semantics, and not the interface semantics. This description includes the pre and post conditions of the operations, and the system uses these operations to automatically generate an interface and the appropriate help [67] HUMANOID s [68, 69, 70] design model captures information about an application s functionality and features of an interface in five dimensions: a) Application semantics design represents the operations and objects that an application program provides, b) Presentation defines the visual appearance of the ....

P.Szekely, P.Luo, and R.Neeches, "Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design," Proc. CHI'92, ACM Conference on Human Factors in Computing Systems, May 1992, pp.507-515.

....transformations. The compositional approach is not new either. Szekely [Szekely 91] models interactive programs based on the notion of communication concepts representing the information that users and programs can communicate, but he focuses only on a class of graphical user interfaces. Humanoid [Szekely 92] applies the compositional approach to creating presentations based on a template based model. Others have modeled other UI components [Card 90, Bleser 90, Myers 90] However, they do not model the overall UI as an integral structure. The most advanced model based UI tools to date are the User ....

....based on a template based model. Others have modeled other UI components [Card 90, Bleser 90, Myers 90] However, they do not model the overall UI as an integral structure. The most advanced model based UI tools to date are the User Interface Design Environment (UIDE) Foley 91, 91a] and Humanoid [Szekely 92] Cartoonist [Sukaviriya 90] which is developed based on the UIDE framework, also belongs here. UIDE does not have a full fledged UI representation model, but the application model is extended with some UI specific details and used as an executable specification. Cartoonist expands the original ....

[Article contains additional citation context not shown here]

Szekely, P., Luo, P., and Neches, R., Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design, in Proceedings of CHI'92. ACM New York, 1992.

....specialized tools to provide assistance during the development process. Szekely 94] Many prototype and UIMS systems (see Section I.D.3) implement the modelbased approach. UIDE [Foley 88] uses design critics which automatically detect and evaluate inconsistencies in the interface design. Humanoid [Szekely 92] uses model editors, model visualizers, interface generators, and alternative generators to create prototypes of a design, possibly before the interface design is complete. It is difficult to distinguish between model based tools and prototyping tools because the model based tools allow such a ....

Szekely, P., P. Luo, and R. Neches. "Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design." Proceedings of SIGCHI'92. 1992: 507-515.

....grows to include actual executable software. Unfortunately, the current generation of tools ignores the earlier phases. Interface builders and other interface programming aids really only help in creating code after the designer has a sense of what is wanted. As we have argued elsewhere [22], although some experimentation is possible, the cost of backing away from a commitment is quite high once much software is built. A great deal is to be gained by maintaining an explicit declarative representation that covers both the design model and the code implementing it. Such a ....

....the model. CARRY OVER OF KNOWLEDGE BETWEEN DESIGN TIME AND RUN TIME TOOLS AND ENVIRONMENTS We have built a number of tools which operate at design time and at run time by making use of the kind of knowledge just listed. ISI s model based user interface development environment is HUMANOID [20, 21, 22]. Its contribution to interface design is that it lets designers express abstract conceptualizations in an executable form, allowing designers to experiment with scenarios and dialogues even before the system model is completely concretized. The consequence is that designers can get an executable ....

[Article contains additional citation context not shown here]

P. Szekely, P. Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In Proceedings of CHI'92, The National Conference on Computer-Human Interaction, May, 1992, pp. 507515.

....means to realize a design; and by preparing tools for interface implementation. With these features, opportunistic design becomes easier to manage. 132 Ping Luo 9.2. 2 Model based Approach Model driven approaches (e.g. UIDE [10] APT [18] Mike [26] SAGE [29] UofA [30] and HUMANOID [35, 36]) use declarative interface models to facilitate interface design with sophisticated services throughout the interface development lifecycle. These models explicitly describe characteristics of interfaces; therefore, by analyzing and annotating the models, the model driven approaches can provide ....

....can be explicitly modeled or can be derived from the semantic models at runtime. Model based systems can be further divided into two categories: automatic interface generation systems (e.g. APT [18] Mike [26] SAGE [29] and UofA [30] and design exploration aids (e.g. UIDE [10] and HUMANOID [35, 36]) Systems that generate interfaces automatically completely insulated designers from detailed design activities. They hide interface design complexity by automating all design decisions (that otherwise would have to be made by human designers) however, they give designers little control over ....

[Article contains additional citation context not shown here]

P. Szekely, P. Luo, and R. Neches. Facilitating the exploration of interface design alternatives: The HUMANOID model of interface design. In Proceedings ACM SIGCHI'92. May 1992, pp. 507-515

No context found.

Szekely 92 P. Szekely, P, Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In Proceedings SIGCHI'92. May 1992, pp. 507-515.

No context found.

P. Szekely, P. Luo, and Neches, R. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In CHI '92: Human Factors in Computing Systems, Monterey, CA, May 1992, pp. 507--515.

No context found.

Szekely, P., Luo, P., Neches, R. Facilitating the exploration of interface design alternatives: the HUMANOID model of interface design. In Conference proceedings on Human factors in computing systems, p. 507-515, 1992.

No context found.

P. Szekely, P, Luo, and R. Neches. Facilitating the Exploration of Interface Design Alternatives: The HUMANOID Model of Interface Design. In Proceedings SIGCHI'92. May 1992, pp. 507-515.

No context found.

Szekely, P., Luo, P., Neches, R. Facilitating the exploration of interface design alternatives: the HUMANOID model of interface design. In Conference proceedings on Human factors in computing systems, p. 507-515, 1992.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC