CHAIN: Developing model-driven contextual help for adaptive user interfaces

• Published in: The Journal of systems and software 2018-01, Vol.135, p.165-190
• Main Author: Akiki, Pierre A.
• Format: Article
• Language: English
• Subjects: Adaptive user interfaces; Contextual help; Design tools and techniques; Model-driven help; Software support
• ISSN: 0164-1212; 1873-1228
• DOI: 10.1016/j.jss.2017.10.017

•CHAIN is an approach for developing model-driven contextual help for adaptive UIs.•CHAIN help models are defined using a language (CHAINXML) and a visual notation.•The definition of help models is supported by Cedar Studio.•CHAIN help models work with both new and legacy application UIs.•Two evaluation studies provided positive indications and insights for future work. Adaptive user interfaces (UIs) change their presentation at runtime to remain usable in different contexts-of-use. Such changes can cause discrepancies between the UI and static help materials, e.g., videos and screenshots, thereby negatively impacting the latter's usefulness (usability and utility). This paper presents Contextual Help for Adaptive INterfaces (CHAIN), which is an approach for developing model-driven contextual help that maintains its usefulness across UI adaptations. This trait is achieved by interpreting the help models at runtime and overlaying instructions on the running adapted UI. A language called Contextual Help for Adaptive INterfaces eXtensible Markup Language (CHAINXML) and a visual notation were developed for expressing and depicting help models. A technique was also devised for presenting CHAIN help models over legacy applications, whether or not their source-code is available. A supporting tool was developed as an extension to Cedar Studio. This work was empirically evaluated in two studies. The first study performed a preliminary evaluation of CHAIN's visual notation. The second study evaluated CHAIN's strengths and shortcomings after using it to develop help for real-life adaptive UIs. The results gave a positive indication about CHAIN's technical qualities and provided insights that could inform future work.