UI Design Pattern Selection Process for the Development of Adaptive Apps Amani Braham University of Sousse, ISITCOM 4011 Hammam Sousse, Tunisia e-mail: amanibraham@gmail.com Félix Buendía Universitat Politècnica de Valencia 46022 Valencia, Spain e-mail: fbuendia@disca.upv.es Maha Khemaja University of Sousse 4000 Sousse, Tunisia e-mail: khemajamaha@gmail.com Faiez Gargouri University of Sfax 3029 Sfax, Tunisia e-mail: faiez.gargouri@isims.usf.tn Abstract—In User Interface (UI) development, UI design patterns constitute a crucial solution that helps to resolve design problems by reusing design knowledge. The diversity of patterns would require deep developer experience to select relevant patterns and would make it difficult to apply the right patterns. This paper proposes an ontology of UI design patterns that enables a potential UI design pattern selection process. We focus particularly on the capability of the Adaptive User Interface Design Pattern (AUIDP) framework in selecting relevant UI design patterns using both ontological and ranking reasoning. This is demonstrated through a service-oriented tool that recommends appropriate patterns. This tool is evaluated with regard to three main factors, including the tool‘s usefulness and practicality, the developed interface quality and the developer productivity. Results show that the tool enhances developer’s accuracy in terms of selecting relevant patterns and hastens the UI development process. Keywords-Adaptive User Interface; Interface specification and design; UI design patterns; Ontology model. I. INTRODUCTION Currently, smartphones and mobile technologies are in the process of an ever-increasing development. The extensive use of mobile devices resulted in a notable increase in the application development industry. This makes the mobile application industry a multi-billion dollar industry [1]. With the increase in the number of mobile applications (a.k.a. apps), developers face a major challenge related to UI development. The statistics presented by Myers et al. [2] reported that the time required for developing user interfaces reaches 50% of the time needed for software development, and their corresponding source code includes 48% of the whole code. These user interfaces are intended to be used by various users with different profiles and needs, and also using different types of devices. A study conducted in [3] showed that 15% of the world’s population has some kind of disability, which could be physical, cognitive, or sensory. The great variety of disabilities that users may be affected by has led to the emergence of adaptive interactive systems [4]. Hence, these systems open up new challenges, as users need adaptive user interfaces that fit their corresponding disabilities and requirements. Therefore, this kind of interface is becoming one of the most dominant part of adaptive systems. However, its development is not a trivial task; it presents a high complexity and takes a long time in such a way that developers often cannot fully cover disabled user’s needs and preferences. Moreover, developing adaptive user interfaces requires a multidisciplinary team with a deep experience in using design knowledge, resolving design problems, as well as choosing the relevant design solution. Within this context, UI design patterns are introduced to support the design of adaptive user interfaces [5], since they attempt to educate designers to build user interfaces [6]. While hundreds of UI design pattern catalogues have been developed and published [7], they tend to be overlooked in practice. The major hurdle in considering these catalogues is how developers can recognize the relevant patterns for solving a specific design problem. This is due to the lack of tools for selecting existing UI design patterns. This might lead to applicability issues that create difficulties for developers to properly select and apply UI design patterns, and makes the design and development of adaptive user interfaces a time consuming and tedious task. Therefore, it becomes mandatory to find an intelligent way to handle, select and use relevant design patterns, to increase the reusability of design knowledge, to decrease the time and complexity of the design and development process and, finally, to improve the quality of adaptive user interfaces for users with disabilities. To tackle the above mentioned challenges, the remainder of this paper presents the Modular UI DEsign Pattern (MIDEP) ontology that enables a potential UI design pattern selection process. This ontology is created using a specific method and augmented with a set of inference rules that provide intelligent support for developers to integrate relevant UI design patterns while developing user interfaces. The selection process is demonstrated through the AUIDP framework, which allows semantic reasoning over the proposed ontology in order to deliver UI design patterns that contribute to the process of developing adaptive mobile applications for users with disabilities. 242 Copyright (c) IARIA, 2020. ISBN: 978-1-61208-761-0 ACHI 2020 : The Thirteenth International Conference on Advances in Computer-Human Interactions