Samuel Lahtinen

Enhancing usability of UML CASE-tools with speech recognition

In the Unified Modeling Language (UML) models are constructed graphically, by drawing diagrams. However, it is not always easy to manipulate diagrams using the CASE-tools of today. For example, some of the functionality and information are hidden behind complex menu or dialog hierarchies, diminishing the usability of the tools. Promoting usability of an existing tool should be done transparently, without disturbing the normal way of using it. Speech recognition can be used as an extra user interface that also allows simultaneous use of other interfaces. UML is a good target for speech recognition since its vocabulary is limited and already specified in the UML metamodel. In this paper, we show how speech recognition can be used to enhance the usage of existing UML CASE-tools. We present the prototype of a speech control system over UML CASE-tools as well as the results of the early usability studies and experiments on the system.

Adding speech recognition support to UML tools

In the Unified Modeling Language (UML) models are constructed graphically, by drawing diagrams. However, it is not always easy to manipulate diagrams using the CASE-tools of today. Typically, functionality and information are hidden behind complex menu or dialog hierarchies, diminishing the usability of the tools. Speech recognition can be used to promote usability of an existing tool as a complementary user interface that allows simultaneous use of other interfaces. In this paper we present an approach to develop speech interfaces to UML tools. We also show that UML is a favorable target to speech recognition and that speech recognition is applicable and mature enough to be used to enhance usage of UML tools. To support our claims, we present a spoken language created for editing UML models, a prototype of a speech control system integrated with Rational Rose, and the results of the early user tests on the system and the language.

Guided Model Creation: A Task-Driven Approach

The abstract syntax of a graphical modeling language is often defined by a metamodel. Typically, a metamodel is given as a class diagram, serving as the grammar of the modeling language. Especially in the case of complex meta-models, there is a need for tool support for assisting the creation of models. In this paper, we propose a method for task-driven creation of models, assuming that the metamodel has been given as a UML class diagram

Service composition for end-users

RESTful services are becoming a popular technology for providing and consuming cloud services. The idea of cloud computing is based on on-demand services and their agile usage. This implies that also personal service compositions and workflows should be supported. Some approaches for RESTful service compositions have been proposed. In practice, such compositions typically present mashup applications, which are composed in an ad-hoc manner. In addition, such approaches and tools are mainly targeted for programmers rather than end-users. In this paper, a user-driven approach for reusable RESTful service compositions is presented. Such compositions can be executed once or they can be configured to be executed repeatedly, for example, to get newest updates from a service once a week.

Decision-making framework for refactoring

Refactoring has been defined as improving code quality without affecting its functionality. When refactoring is overlooked in daily development, the likelihood of larger refactorings increases with time. Disadvantages of larger refactorings include that they disrupt the daily work, require additional planning effort, and often they need to be justified to stakeholders. In this paper, we investigate with interviews how professionals make refactoring decisions. As a result, we present a framework for decision making for larger refactoring operations describing the key stages in a refactoring workflow. Furthermore, one actual industry case of refactoring decision making is presented in detail.

A Specification Technique for Model Based Derivation of Speech Interfaces

We introduce a specification technique that uses a combination of statechart templates and structural models for deriving spoken command languages for the domains specified by the structural models. We apply the technique for the domain of UML editing

Hammer and Nails - Crucial Practices and Tools in Ad Hoc Student Teams

We have observed students teams on a software engineering project course to understand what software engineering practices they end up using and how do they experience the usefulness of the selected practices and tools. In our context, the most often applied practices and tools were planning meeting, commitment to using tasks, self-selected communication tools, a revision control, and project management systems. We found out that student valued various face-to-face activities the most - even more than they were able to practice them. Finally, the wide variety of communication tools students take even from their leisure time and apply in the course setting surprised us.

Unwasted DASE: Lean Architecture Evaluation

A software architecture evaluation is a way to assess the quality of the technical design of a product. It is also a prime opportunity to discuss the business goals of the product and how the design bears on them. But architecture evaluation methods are seen as hard to learn and costly to use. We present DASE, a compact approach that combines carefully selected key parts of two existing architecture evaluation methods while making evaluation lean and fast. We have applied DASE in three industrial cases and the early results show that even a one-day evaluation workshop yields valuable results at a modest cost.

Refactoring Patterns, Practices for Daily Work

In this paper, we describe patterns that can be used to find ways to integrate refactoring into everyday work in a software project. They are a part of a larger refactoring patterns pattern collection. We introduce the pattern collection and the background and describe three patterns in detail. Revision control logging pattern helps to maintain the rationale of the refactoring operations and separates refactoring from, for instance, feature development and bug fixes. Embed small refactorings encourages you to do the minor tweaks, fixes, and changes during your daily routines. Finally, Establish safety net guides you to setup unit tests and revision control before starting refactoring.

To UML or not to UML?: Empirical Study on the Approachability of Software Architecture Diagrams

Software architecture design is key to building systems that meet quality demands. Choosing the appropriate way to model the architecture ensures it is rightly understood by everyone involved. UML diagrams are commonly used in software engineering but free-form diagrams are almost as common. In this paper, we study the factors influencing the approachability of diagrams, and particularly whether there is a difference in the approachability between UML and non-UML diagrams and colored and black & white diagrams. Our results show that colors do not necessarily increase the approachability of diagrams and free-form diagrams can suffer from ambiguousness. We conclude that simplicity and correctness are key factors when modeling architectures.