Kinch, Martin ORCID: 0000-0002-4737-7118 and Keates, Simeon ORCID: 0000-0002-2826-672X (2021) Contextual cues: the role of machine learning in supporting contextually impaired users. In: Universal Access in Human-Computer Interaction. Design Methods and User Experience 15th International Conference, UAHCI 2021, Held as Part of the 23rd HCI International Conference, HCII 2021, Virtual Event, July 24–29, 2021, Proceedings, Part I. Lecture Notes in Computer Science - Information Systems and Applications, incl. Internet/Web, and HCI (LNISA), 12768 . Springer, Cham, pp. 567-569. ISBN 978-3030780920; 978-3030780913 (doi:https://doi.org/10.1007/978-3-030-78092-0_39)

PDF (AAM) 36744_KINCH_Contextual_cues.pdf - Accepted Version Restricted to Repository staff only Download (192kB) | Request a copy

Abstract

This paper explores the theoretical aspect of providing context to contextually impaired individuals and offers some considerations on how a machine learning system can be adapted to learn contextual clues and then provide these to users. Context itself, is the clarifying component of a situation and helps people to understand what is happening and why. Some people struggle to understand the situations that they are in, due to several issues, particularly those with sensory or social impairments, such as autism. These impairments can be interpreted as form of contextual blindness, where a lack of awareness of certain contextual clues render communication or under-standing of it difficult. These clues can be as simple as a person’s current lo-cation, to the more nuanced impact of a conversational partner’s body language. Since awareness of the current context of a situation can help to clarify its meaning, it is expected that provision of contextual information to con-textually impaired individuals will be beneficial; by giving them a point of reference for the tone of the situation they are in and supporting them to re-act accordingly. The results presented demonstrate the impact of contextual information in a specific social situation (dating), giving an indication of the benefits of both context and contextual awareness, which provide the basis for further investigation.

Item Type:	Conference Proceedings
Title of Proceedings:	Universal Access in Human-Computer Interaction. Design Methods and User Experience 15th International Conference, UAHCI 2021, Held as Part of the 23rd HCI International Conference, HCII 2021, Virtual Event, July 24–29, 2021, Proceedings, Part I
Uncontrolled Keywords:	context; machine learning; contextual support; autism support; contextual information processing
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science R Medicine > RE Ophthalmology
Faculty / School / Research Centre / Research Group:	Faculty of Engineering & Science
Last Modified:	04 Jul 2022 13:00
URI:	http://gala.gre.ac.uk/id/eprint/36744

Actions (login required)

View Item

Downloads