Software Engineering

Model-driven engineering (MDE), including Model Based approaches of user interfaces describes the User interface and its related aspects (e.g. tasks, domain, context of use) using set of models that can derive a final user interface. Model-Based user interface approaches give the researchers a new designing methodology that eases the creation of user interface and tackles the problems of producing a new design for every new device and every modality (graphical, vocal, gesture, …)
The design of interactive systems for an ambient intelligent environment (AMI) poses many challenges due to the diverse number of devices, and interaction modalities available in the environment together with restrictions imposed by making the interactive system usable by people with disabilities. This work proposes a framework that integrates model based techniques with AMI to build an applicable solution which can allow the designers to design and develop multi-device and multi-modal user interfaces to fit the emergent needs of people with disabilities in an AMI environment through a number of model-transformations.

Testing GUIs for correctness can enhance the entire system's safety, robustness, and usability. The major difficulty in testing a GUI comes from fact that it is impossible to cover all possible interactions with all possible paths they might lead to. This paper introduces a model-based testing framework of GUIs that can automatically generate test cases from the task and dialog models.

The wide variety of interactive devices and modalities an interactive system must support has created a big challenge in designing a multi-platform user interface and poses a number of issues for the design cycle of interactive systems. Model-Based User Interface Design (MBUID) ...

Background: It is widely recognized that software effort estimation is a regression problem. Model Tree (MT) is one of the Machine Learning based regression techniques that is useful for software effort estimation, but as other machine learning algorithms, the MT has a large space of configurations and requires to carefully setting its parameters. The choice of such parameters is a dataset dependent so no general guideline can govern this process which forms the motivation of this work. Aims: This study investigates the effect of using the most recent optimization algorithm called Bees algorithm to specify the optimal choice of MT parameters that fit a specific dataset and therefore improve prediction accuracy. Method: We used MT with optimal parameters identified by the Bees algorithm to construct software effort estimation model. The model has been validated over eight datasets come from two main sources: PROMISE and ISBSG. Also we used 3-Fold cross validation to empirically assess the prediction accuracies of different estimation models. As benchmark, results are also compared to those obtained with Stepwise Regression, Case-Based Reasoning and Multi-Layer Perceptron. Results: The results obtained from combination of MT and Bees algorithm are encouraging and outperforms other well-known estimation methods applied on employed datasets. They are also interesting enough to suggest the effectiveness of MT among the techniques that are suitable for effort estimation. Conclusions: The use of the Bees algorithm enabled us to automatically find optimal MT parameters that are required to construct accurate effort estimation model for each individual dataset.

One of the major problems with software project management is the difficulty to predict accurately the required effort for developing software applications. Analogy Software effort estimation appears well suited to model problems of this nature. The analogy approach may be viewed as a systematic development of the expert opinion through experience learning and exposure to analogue case studies. The accuracy of such model depends on characteristics of datasets. This paper examines the impact of feature subset selection algorithms on improving the accuracy of analogy software effort estimation model. We proposed a feature subset selection algorithm based on fuzzy logic for analogy software effort estimation models. Validation using two established datasets (ISBSG, Desharnais) shows that using fuzzy features subset selection algorithm in analogy software effort estimation contribute to significant results as other algorithms: Hill climbing, Forward subset selection, and backward subset selection do.

Delivering accurate software effort estimation has been a research challenge for a long time, where none of the existing estimation methods has proven to consistently deliver an accurate estimate. Previous studies have demonstrated that estimation by analogy (EBA) is a viable alternative to other conventional estimation methods in terms of predictive accuracy. EBA offers a way to use a formal method with data from a past project to derive a new estimate. Two important research areas in EBA are addressed in this paper: software projects similarity measurement and attribute weighting. However, the inherent uncertainty of attribute measurement makes similarity measurement between two software projects subject to considerable uncertainty. To tolerate such inherent uncertainty we propose a new similarity measurement method by combining the advantages of Fuzzy Set Theory and Grey Relational Analysis. In addition, since each attribute has different influence on the project retrieval we propose a new approach to deal with this issue based upon the idea of Kendall's coefficient of concordance between the similarity matrix of project attributes and the similarity matrix of known effort values of the dataset. Our results show improved prediction accuracy when multiple project attributes are used with determined weights.

Relaying on early effort estimation to predict the required number of resources is not often sufficient, and could lead to under or over estimation. Software Project managers may not recognize that software development process should be refined regularly and that software prediction made at early stage of software development is yet kind of guesses. Even good predictions are not sufficient with inherent uncertainty and risks. The stage-effort estimation allows project manager to re-allocate correct number of resources, re-schedule project and control project progress to finish on time and within budget. In this paper we propose an approach to utilize prior effort records to predict stage effort. The proposed model combines concepts of Fuzzy set theory and association rule mining. The results were good in terms of prediction accuracy and have potential to deliver good stage-effort estimation.

A reliable and accurate similarity measurement between two software projects has always been a challenge for analogy-based software cost estimation. Since the effort for a new project is retrieved from similar historical projects, it is essentially to use the appropriate similarity measure that finds those close projects which in turn increases the estimation accuracy. In software engineering literature, there is a relatively little research addressed the issue of how to find out similarity between two software projects when they are described by numerical and categorical features. Despite simplicity of exiting similarity techniques such as: Euclidean distance, weighted Euclidean distance and maximum distance, it is hard to deal with categorical features. In this paper we present two approaches to measure similarity between two software projects based on fuzzy C-means clustering and fuzzy logic. The new approaches are suitable for both numerical and categorical features.

Context: Software effort estimation at early stage is a crucial task for project bedding and feasibility study. Since collected data at early stage of software development lifecycle is always imprecise and uncertain, it is very hard to deliver accurate estimate. Analogy-based estimation, which is one of the popular estimation methods, is rarely used at early stage because of uncertainty associated with attribute measurement and data availability.

Analogy estimation is a well known approach for software effort estimation. The underlying assumption of this approach is the more similar the software project description attributes are, the more similar the software project effort is. The most difficult activity in analogy estimation is how to derive a new estimate from retrieved solutions. Using retrieved solutions without adjustment to considered problem environment is not often sufficient. Thus, they need some adjustment to minimize variation between current case and retrieved cases. The main objective of the present paper is to investigate the applicability of fuzzy logic based software projects similarity measure to adjust analogy estimation and derive a new estimate. We proposed adaptation techniques which take into account the similarity between two software projects in terms of each feature. In earlier work, we proposed and validated a similarity measure between software projects based on fuzzy C-means. This similarity measure will be guided towards deriving new estimate.

Background: Adaptation technique is a crucial task for analogy based estimation. Current adaptation techniques often use linear size or linear similarity adjustment mechanisms which are often not suitable for datasets that have complex structure with many categorical attributes. Furthermore, the use of nonlinear adaptation technique such as neural network and genetic algorithms needs many user interactions and parameters optimization for configuring them (such as network model, number of neurons, activation functions, training functions, mutation, selection, crossover, etc.). Aims: In response to the abovementioned challenges, the present paper proposes a new adaptation strategy using Model Tree based attribute distance to adjust estimation by analogy and derive new estimates. Using Model Tree has an advantage to deal with categorical attributes, minimize user interaction and improve efficiency of model learning through classification. Method: Seven well known datasets have been used with 3-Fold cross validation to empirically validate the proposed approach. The proposed method has been investigated using various K analogies from 1 to 3. Results: Experimental results showed that the proposed approach produced better results when compared with those obtained by using estimation by analogy based linear size adaptation, linear similarity adaptation, 'regression towards the mean' and null adaptation. Conclusions: Model Tree could form a useful extension for estimation by analogy especially for complex data sets with large number of categorical attributes.