The evolution of autonomous vehicles (AVs) requires a paradigm shift towards the integration of human factors to improve safety and efficiency at levels 2,3 and 4 of automation. This paper presents a comparison of three different fusion technologies (Low-Level fusion, Medium-Level fusion, and a hybrid fusion), highlighting the critical role of multimodal data integration and semi-supervised learning in predicting and adapting to levels of driver awareness. Our approach uses semi-supervised learning to deal with the data labelling problem, using unlabelled data to train an autoencoder and sparsely labelled data to train a 4-state classifier. Our model facilitates the fusion of data from different physiological signals, including skin electrodermal activity, heart rate, body temperature and acceleration. Using real driving data, the Medium-Level fusion approach gives the best performance, achieving 84% accuracy in predicting situations where the user may not be aware enough to take control of the vehicle. This research highlights the essential nature of fusion technologies to create adaptive and user-centred AV systems.

The relation between affect-driven feedback and engagement on a given task has been largely investigated. This relation can be used to make personalised instructional decisions and/or modify the affect content within the feedback. However, although it is generally assumed that providing encouraging feedback to students should help them adopt a state of flow, there are instances where those messages might result counterproductive. In this paper, we present a case study with 48 secondary school students using an Intelligent Tutoring System for arithmetical word problem solving. This system, which makes some common assumptions on how to relate affective state with performance, takes into account subjective (user's affective state) and objective information (previous problem performance) to decide the upcoming difficulty levels and the type of affective feedback to be delivered. Surprisingly, results revealed that feedback was more effective when no emotional content was included, and lead to the conclusion that purely instructional and concise help messages are more important than the emotional reinforcement contained therein. This finding shows that this is still an open issue. Different settings present different constraints generating related compounding factors that affect obtained results. This research confirms that new approaches are required to determine when, how and where affect-driven feedback is needed. Affect-driven feedback, engagement and their mutual relation have been largely investigated. Student's interactions combined with their emotional state can be used to make personalised instructional decisions and/or modify the affect content within the feedback, aiming to entice engagement on the task. However, although it is generally assumed that providing encouraging feedback to the students should help them adopt a state of flow, there are instances where those encouraging messages might result counterproductive. In this paper, we analyze these issues in terms of a case study with 48 secondary school students using an Intelligent Tutoring System for arithmetical word problem solving. This system, which makes some common assumptions on how to relate affective state with performance, takes into account subjective (user's affective state) and objective (previous problem performance) information to decide the difficulty level of the next exercise and the type of affective feedback to be delivered. Surprisingly, findings revealed that feedback was more effective when no emotional content was included in the messages, and lead to the conclusion that purely instructional and concise help messages are more important than the emotional reinforcement contained therein. This finding, which coincides with related work, shows that this is still an open issue. Different settings present different constraints and there are related compounding factors that affect obtained results, such as the message's contents and their target, how to measure the effect of the message on engagement through affective variables considering other issues involved, and to what extent engagement can be manipulated solely in terms of affective feedback. The contribution here is that this research confirms that new approaches are needed to determine when, how and where affect-driven feedback is needed. In particular, based on our previous experience in developing educational recommender systems, we suggest the combination of user-centred design methodologies with data mining methods to yield a more effective feedback.

Abstract Higher and further education providers are facing the challenge of supporting the interaction needs of an increasing number of students who feature accessibility preferences to use both elearning contents and services. In the next future, we can expect that, within adaptive elearning systems, both automatic and manual procedures will interoperate to elicit users' interaction needs for ensuring accessibility. In this paper, we report findings from the user experience with the self‐assessment of interaction needs, as part of a content personalization system, which tackles possible mismatches in the interaction between the user and the learning objects. All stakeholders involved in providing this service along with intended user groups (students with visual, auditory or mobility impairments, and without impairments) participated in the evaluation with over 100 users described in this paper. From the evaluation, results follow that our approach allows students to self‐assess and report adequately their interaction preferences. Furthermore, the paper describes findings of interest and open issues about how massive online courses may address the accessibility needs of an increasing number of elearning users.

Abstract Collaborative indicators derived from participants' interactions can be used to support and improve their collaborative behaviour. In this research, we focus on automatically identifying recommendation opportunities in the Collaborative Logical Framework from participants' interactions. Different information sources have been considered: (a) statistical collaborative indicators; (b) social interactions; (c) opinions received by the participants via ratings; and (d) users' affective state and personality. The recommendations have been elicited considering the generality and transferability of the participants' interactions provided by the Collaborative Logical Framework. As a result, three scenarios have been identified that lead us to propose meaningful grouping suggestions and recommendations, which ultimately aimed to ground an informed personalized support to the participants in intensive collaborative frameworks.

Abstract Social online learning environments provide new recommendation opportunities to meet users' needs. However, current educational recommender systems do not usually take advantage of these opportunities. To progress on this issue, we have proposed a knowledge engineering approach based on human–computer interaction (i.e. user‐centred design as defined by the standard ISO 9241‐210:2010) and artificial intelligence techniques (i.e. data mining) that involve educators in the process of eliciting educational oriented recommendations. To date, this approach differs from most recommenders in education in focusing on identifying relevant actions to be recommended on e‐learning services from a user‐centric perspective, thus widening the range of recommendation types. This approach has been used to identify 32 recommendations that consider several types of actions, which focus on promoting active participation of learners and on strengthening the sharing of experiences among peers through the usage of the social services provided by the learning environment. The paper describes where data mining techniques have been applied to complement the user‐centred design methods to produce social oriented recommendations in online learning environments.

The emotional situation of the learner can influence the learning process. For this reason, we are researching how educational recommender systems can take advantage of affective computing to improve the recommendation support in educational scenarios. The paper reports works carried out involving 18 educators and 77 learners to elicit and design emotional feedback to be provided for learners in terms of personalized recommendations. To this end, user centered design methods and data mining techniques are used.

In this paper we propose a series of indicators, which derive from user's interactions with mouse and keyboard. The goal is to evaluate their use in identifying affective states and behavior changes in an e-learning platform by means of non-intrusive and low cost methods. The approach we have followed study user's interactions regardless of the task being performed and its presentation, aiming at finding a solution applicable in any domain. In particular, mouse movements and clicks, as well as keystrokes were recorded during a math problem solving activity where users involved in the experiment had not only to score their degree of valence (i.e., pleasure versus displeasure) and arousal (i.e., high activation versus low activation) of their affective states after each problem by using the Self-Assessment-Manikin scale, but also type a description of their own feelings. By using that affective labeling, we evaluated the information provided by these different indicators processed from the original user's interactions logs. In total, we computed 42 keyboard indicators and 96 mouse indicators.

In this paper we address an open key issue during the development of web-based educational systems. In particular, we provide an educational-oriented approach for building personalised e-learning environments that focuses on putting the learners' needs in the centre of the development process. Our approach proposes user centred design methodologies involving interdisciplinary teams of software developers and domain experts. It is illustrated in an adaptive e-learning system, where a MOOC (Massive Open Online Course) was taken by nearly 400 learners. In particular, we report where user centred design methods can be applied along the e-learning life cycle to designing and evaluating personalisation support through recommendations in learning management systems.

Willow is a free-text Adaptive Computer Assisted Assessment system, which supports natural language processing and user modeling. In this paper we discuss the benefits coming from extending Willow with recommendations. The approach combines human computer interaction methods to elicit the recommendations with data mining techniques to adjust their definition. Following a scenario-based approach, 12 recommendations were designed and delivered in a large scale evaluation with 377 learners. A statistically significant positive impact was found on indicators dealing with the engagement in the course, the learning effectiveness and efficiency, as well as the knowledge acquisition. We present the overall system functionality, the interaction among the different subsystems involved and some evaluation findings.

When building a knowledge-based recommender along the eLearning life cycle, the following issues have to be considered: a) the user interface design of the tools required, b) the process to design/generate the recommendations, c) the process to select the appropriate recommendations, and d) the management of the users' interactions. We are defining a user-centered evaluation approach that copes with those issues and drives the recommender building process in three consecutive steps: 1) elicitation of pedagogically sound recommendations validated by users with a collaborative review, 2) acquisition and validation of the user features to select the appropriate recommendations for the current context, and 3) analysis of the recommendations provided and evaluation of their impact on the user.

When students are responsible for their own collaborative learning process, they need information on collaboration to manage the process. We note that collaborative environments fail to provide users with information on the collaboration process. This research proposes an open collaborative learning experience design that provides learners with an environment where they can manage the collaboration. The design provides information on collaboration to improve management. An inferring method, which uses statistical indicators of student interaction in forums and data mining technology, works to obtain information on student collaboration and groups students according to their level of collaboration. This information will be shown to students and tutors to improve the learning process and its management.

Active modeling is required in learning settings to cope with the dynamic evolution of the knowledge, since learners competences evolves over time as they participate in the course activities. Moreover, one of the main issues in a competence based eLearning process is to deliver personalized instructional designs adjusted to both 1) intrinsic characteristics of users (i.e. learning styles) and 2) the desired and achieved competences in the learning process (i.e. specific and generic competences). This delivery includes the adaptation of the content and the activities in a learning scenario based on a dynamic user model that evolves according to user interactions. In this paper, an approach to support Conditional Plans Generation (IMS Learning Designs) in the context of a virtual learning environment is presented. The process is supported by a pervasive usage of standards and specifications (IMS family of specifications) in conjunction with an integral user modeling.

This paper describes our approach to develop Computer Adaptive Testing functionality for different Learning Management Systems following educational standards and specifications (defined by the IMSConsortium). This functionality can be integrated into standard-based personalized learning routes and is being developed as a Service Oriented Architecture(SOA) to support its generality and applicability.

Education and work are inextricably linked to the peoplepsilas lives in our modern society and they are the main agents for a successful social integration, but at the same time they can become barriers for the inclusion of the disabled. The technology can be used to overcome these barriers offering an adaptive layer that can cope with the needs of the people with disabilities. Our current research is focused on building user models and generating dynamic recommendations to provide personalized e-learning and e-mentoring. The research is applied in the context of communities for health and independent living.

In this paper we present research works we are addressing in EU4ALL project (IST-2006-034778) to enable Higher Education (HE) institutions to support and attend the accessibility needs of their students. This approach is based on integrating learning and management of the learning in terms of workflows to support the different types of existing scenarios with a twofold objective. First, involving non-technical staff in their definition. Second, using standard-based learning management systems (LMS). A combination of design and runtime adaptations through IMS Learning Design (IMS-LD) specification is being used, following the aLFanet approach (IST-2001-33288).

The technology is expected to attend the learning needs of the students in a personalized and inclusive way. Our approach relies on combining design and runtime adaptations for all with a pervasive use of standards and specifications. However, there is currently insufficient support provided by the specifications. For this reason, we propose a multi-agent architecture to produce dynamic recommendations that complements limitations of the universal design approach. Evaluations with users justify the need of this support.