A systematic literature review on intelligent user interfaces: preliminary results

Reviewer: Franz J Kurfess

Designers of user interfaces often face a fundamental dilemma: how much of the underlying functionality of the system should they expose to the user An experienced user may want direct access to most, or all, of the functionality, while casual users would benefit from a more minimalist approach. In principle, the solution for this dilemma is smart user interfaces (UIs), also referred to as intelligent, adaptive, or context-sensitive UIs. This paper presents initial results from a systematic literature review of intelligent user interfaces (IUIs). They define a research protocol for the review, with a clearly specified objective, research questions, specific search strings, the search engines used, and inclusion/exclusion criteria. Out of an initial set of 561 papers retrieved from two scientific search engines (Scopus and Web of Science) in response to an elaborate query, after the application of the inclusion/exclusion criteria, 151 were selected for full reading. In the results analysis, the authors extract relevant data guided by the criteria identified in the research questions. Most papers were published in conferences (127) versus journals (24); books like research monographs and edited collections apparently didn't make the cut [1,2,3,4]. Not surprisingly, "intelligent" and "adaptive" are the most frequent modifiers combined with user interface, and a wide range of models and software technologies are reported. One of the prominent application domains is healthcare. Further analysis will be performed based on a grounded theory methodology. I'm looking forward to a more detailed discussion of this topic in a full report.