Home // CENTRIC 2014, The Seventh International Conference on Advances in Human-oriented and Personalized Mechanisms, Technologies, and Services // View article
User-centric Recommendation Model for AAC based on Multi-criteria Planning
Authors:
Sebastián Sastoque Hernández
Daniela Mancilla Valdés
Jesús Mendoza Medina
Marcela Iregui Guerrero
Keywords: Personalized recommendation, AAC, Multi-criteria, Planning, Hybrid recommender, Phrase Prediction.
Abstract:
Recommendation algorithms are useful to fast decision-making in complex scenarios. They are based in grasping knowledge from the user and his environment to personalize applications according to a specific domain. However, the complexity of each domain and the dynamic relations between different items and criteria can be challenging. In Alternative and Augmentative Communication (AAC), assistive technologies have been envisaged to help people with neuro-linguistic disorders in stories construction. Nevertheless, there are still many problems to overcome, related to the algorithms and their application. In this work, we present an original approach based on knowledge domain modelling, user archetypes, learning strategies and multi-criteria planning to assist the user in managing a set of decisions. The proposed algorithm was applied in the context of AAC systems, aimed to help aphasic persons in the construction of coherent and semantically correct stories about their daily activities. The system was validated by an aphasic person and several healthy individuals, in terms of the precision and efficiency of the method, compared with other approaches. The results demonstrate that the presented method has a good precision in the recommendation, allows faster decisions and is well adapted to AAC and can be extended to similar applications.
Pages: 66 to 72
Copyright: Copyright (c) IARIA, 2014
Publication date: October 12, 2014
Published in: conference
ISSN: 2308-3492
ISBN: 978-1-61208-369-8
Location: Nice, France
Dates: from October 12, 2014 to October 16, 2014