{{Short description|User interface type}} {{redirect|Interface agent|an intelligent agent that interacts with its environment|Embodied agent}} An '''intelligent user interface''' ('''intelligent UI''', '''IUI''', or sometimes '''interface agent''') is a [[user interface]] (UI) that involves some aspect of [[artificial intelligence]] (AI or computational intelligence). There are many modern examples of IUIs, the most famous (or infamous) being the Microsoft [[Office Assistant]], whose most recognizable agentive representation was called "[[Clippy]]".
Generally, an IUI involves the computer-side having sophisticated knowledge of the domain and/or a model of the user. These allow the interface to better understand the user's needs and personalize or guide the interaction.
== History == Probably the earliest examples of what could be considered true IUIs appeared in the [[Intelligent tutoring system|Intelligent Computer Assisted Instruction]] (ICAI, aka. [[intelligent tutoring system]]s) community, which arose in the 1960s and 1970s<ref name="BurtonBrown1979">{{cite journal|last1=Burton|first1=Richard R.|last2=Brown|first2=John Seely|title=An investigation of computer coaching for informal learning activities|journal=International Journal of Man-Machine Studies|volume=11|issue=1|year=1979|pages=5–24|issn=0020-7373|doi=10.1016/S0020-7373(79)80003-6|url=http://www.dtic.mil/get-tr-doc/pdf?AD=ADA061268|archive-url=https://web.archive.org/web/20170924160749/http://www.dtic.mil/get-tr-doc/pdf?AD=ADA061268|url-status=dead|archive-date=September 24, 2017|url-access=subscription}}</ref> and become popular (among academics) in the 1980s. Also, in the early 1980s, as [[expert system]]s took hold in the AI community, expert systems were applied to UIs (e.g., the aptly-named "WIZARD" system<ref name="Shrager92">{{cite conference |last1=Shrager|first1=Jeff|first2=Tim|last2=Finin|title=An expert system that volunteers advice|book-title=Proc. 2nd Annual National Conf. on Artificial Intelligence (AAAI-82)|place=Pittsburgh, PA|pages=339–340|date=1982|url=https://www.aaai.org/Papers/AAAI/1982/AAAI82-080.pdf}}</ref>). In the 1990s the application of [[plan inference]] to interaction formed the basis for research<ref>{{cite journal |last1=Cawsey |first1=Alison |title=User modelling in interactive explanations. User Model User-Adap Inter 3, 221–247 (1993)|journal=User Modeling and User-Adapted Interaction|date=1993 |volume=3|doi=10.1007/BF01257890 |s2cid=5555505 |url=https://doi.org/10.1007/BF01257890 |access-date=4 January 2023|url-access=subscription }}</ref> in what then was named [[Natural user interface|natural interfaces]] (the term has later come to evolve to mean full-body interaction). Later IUIs, such as [[Clippy]], are more statistically-based, using [[machine learning]] methods to decide how to tune the interactive experience to the individual user. In the 2000s this strand of research often is labeled [[personalization]], most often employing various [[recommender system]] techniques to adapt the behavior of an interface or an entire interactive system to individual user preferences.
== Definitional difficulty == What constitutes "intelligent" is potentially disputable, as is what counts as an "interface". The field is in practice defined by the community of researchers and the channels they publish in.
== Research == Research in intelligent user interfaces is published in general Human-Computer Interaction conferences and journals such as [[Conference on Human Factors in Computing Systems|CHI]] or [[UIST (conference)|UIST]] as well as in some artificial intelligence research channels such as those hosted by the [[AAAI]], but most importantly there are the dedicated conference series on [https://dl.acm.org/conference/iui Intelligent User Interfaces] (since 1988), [https://dl.acm.org/conference/recsys Recommender Systems] (since 2007), and [https://dl.acm.org/conference/umap User Modeling, Adaptation and Personalization] as well as the journal [https://www.springer.com/journal/11257 User Modeling and User-Adapted Interaction] (since 1990).
== Humanizing cues == As practical concept of [[social interface design]], social interface is seen in the studies of [[human-computer interaction]] (in particular, its [[computer interface]] aspect). The basic thesis is that where a computer interface is more akin to another human, it can facilitate correct responses from users during human-to-computer interaction. Software that can provide such humanizing cues often does it by creating interface with human-like quality (such as giving recognizable [[gender]] to a [[software agent]]).<ref>Tourangeau R.; Couper M.P.; Steiger D.M., '' Humanizing self-administered surveys: experiments on social presence in web and IVR surveys'', Computers in Human Behavior, Volume 19, Number 1, January 2003, pp. 1-24(24), [http://www.ingentaconnect.com/content/els/07475632/2003/00000019/00000001/art00032]</ref> Studies are often concerned with how should such agents (like the [[Microsoft Agent]]) be designed to make them more appealing (is having [[facial expression]]s efficient, should the agent be [[anthropomorphic]], and so on).<ref>Michael A. Cusumano, Richard W. Selby, ''Microsoft Secrets: How the World's Most Powerful Software Company Creates Technology, Shakes Markets and Manages People''. Simon and Schuster, 1998, {{ISBN|0-684-85531-3}}, [https://books.google.com/books?id=GixEgGs5qXcC&dq=%22Social+interface%22&pg=PA178 Google Print, p.178]</ref>
==See also== * [[Natural language user interface]] ** [[Chatterbot]] * [[Recommender system]]
== References == {{Reflist}}
== Links ==
* [http://www.aaai.org/Conferences/AAAI/aaai.php AAAI Conference] * [https://dl.acm.org/conference/iui International Conference on Intelligent User Interfaces] * [https://dl.acm.org/conference/recsys ACM Conference On Recommender Systems] * [https://www.springer.com/journal/11257 Journal on User Modeling and User-Adapted Interaction]
[[Category:User interfaces]]