Smart Learning Environments aim at automatically adapting the learning experience based on learner's context. When this context is not restricted to formal settings, SLEs are a promising solution for automatically connecting formal education with informal learning opportunities that emerge in di erent physical and virtual spaces. To achieve this, SLEs can bene t from both the information from the formal learning design as well as the capability of sensing and analyzing the progress of each learner. In previous research, we have devised an architecture to interconnect the di erent technologies that form an SLE capable of connecting formal and informal learning across-spaces. This paper goes a step forward by exploring the information ow needed to model the current context and state of the learner to eventually trigger informal learning interventions.
Smart Learning Environments (SLEs) seek to automatically provide personalized
support to the students considering their individual needs and context [
When the learning context is not restricted to formal settings, SLEs are a
promising solution for connecting formal education with informal learning [
Prior attempts to bridge both types of learning were related to self-regulated
learning, where learners could establish communities of learning related with
their concerns [
In [
The rest of the paper is structured as follows. Section 2 exempli es the support of SLEs in a sample scenario- Section 3 describes the information ow happening in an SLE to support that scenario and how the learning design inuences the di erent stages. Finally, section 4 present the main conclusions and future work. 2
For the sake of illustration, we present a scenario that re ects how an SLE can achieve such connect. Stella teaches a Natural Science course in a high school. She is devoting the following three weeks to teach about the fauna, ora and landscape in the local region, focusing on one of these topics each week. With this purpose, the teacher prepares di erent activities for her students. First, she gives a short presentation in class of relevant aspects in one of the aforementioned domains. Then, she proposes the learners to do a reading or watch a video on the subject available in the VLE at home. The following day, in class she organizes a debate among the students discussing about the revised material. Finally, she asks her students to ll out a progress quiz through the VLE by the end of the week. This sequence of formal learning activities is described in the learning design.
In previous years, Stella observed that her students have some issues to re ect the theoretical concepts in the real world, so she decides to support her lessons with an SLE. As the SLE needs to notify students about possible activities, she suggests her students to install a companion app in their personal mobile devices, and eventually provide an informed consent regarding the use of the data it collects for academic purposes.
After the rst week, Pedro has attended all in-class activities (as reported by Stella in the assistance report), watched the mandatory video but he did not checked any of the available readings (as reported by the VLE). As well, he did not score very high in the progress quiz (as reported again by the VLE). The SLE consults the learning design to seek the data sources linked to the di erent activities and gather the information about the performance of the students. As a result, from the information above, the SLE detected that Pedro's knowledge in local fauna is low. Pedro decided to install the companion app and, one day, he happens to walk in his town near a natural park, as reported by the app. The SLE, with the information coming from the app, detects that it is a suitable context to support Pedro, so it proposes Pedro an informal activity to identify the di erent tree species in the park, considering the topic of the learning situation. 3
In order to support the scenario described above, the SLE has to manage di erent
information related with the actions of the students during the learning situation.
The information ow that takes places in SLEs can adhere to the
sense-analyzereact model [
The inclusion of the learning design in the information ow can bene t SLEs
to better orchestrate the aforementioned actions. Through speci cations such as
IMS-LD [
This ow begins with the deployment of the teacher's learning design among the learning spaces where the speci ed activities and resources take place. The information contained in the learning design (such as activity timing, topics, goals and related learning spaces) will be used during the next phases of the information ow.
Once the learning design has been deployed, the SLE can proceed with the collection of data related with the student participation in the learning situation. Each learning space o ers measuring tools to track how students are participating in the di erent activities or how they interact with the available resource (e.g., timestamped actions in the VLE such as retrieving a document or answering a quiz, or presence and location in the physical space). The SLE gathers all the appropriate information from the pertaining learning spaces, according to the learning design, and prepares it for its analysis.
In the analysis phase, the SLE extracts the indicators that complete the actionable information the SLE has about the student. This information can be classi ed in two sets: (1) the student's model, that contains information derived from the progress and performance through the di erent activities (e.g. degree of knowledge of a certain topic); and (2) the student's context, more related to the current conditions of the learner (e.g. his location and whether he is currently connected to the SLE). Both of these sets are constantly evolving as the learning situation takes place. During this phase, the SLE relies on the learning design to determine the appropriate analysis to perform, as well as to match the student's model with the goals and topics considered in the design.
With the student's model in continuous evolution, the SLE can automatically react and intervene and provide the personalized support to the students. To do so, the SLE evaluates previously con gured reaction rules and activity templates, leading to a personalized, informal support. The reaction rules trigger by evaluating information from the student model that concerns the topics and goals of the situation. On the other hand, the activity templates contain a variety of possible reactions, from simply informing the teacher (so that she can decide how to react) to suggesting simple or more complex activities (e.g. a personalized reading or quiz, to take pictures and comment a physical resource, or to identify the tree species as in the scenario). It should be noted that rules and templates are de ned before in a general fashion, but are tailored on the y to the corresponding topics, goals and context. All this reaction process takes place automatically without the teacher's intervention.
SLEs present potential opportunities to enhance students' learning experiences by connecting formal and informal learning. To this end, these environments should properly align their understanding of the students with the learning goals of the formal education. This paper presents how the inclusion of the learning design in the information ow of an SLE can help in the construction of the model of the student and in the provision of appropriate resources. Nevertheless, there is pending work in the automatic provision by SLEs of appropriate resources according to the learning design. In future work, the authors will collaborate with stakeholder for the de nition of the reaction templates and the provision of resources to present to the students in di erent learning spaces and context.
This research is partially funded by the European Regional Development Fund and the National Research Agency of the Spanish Ministry of Science, Innovations and Universities under project grants TIN2017-85179-C3-2-R, by the European Regional Development Fund and the Regional Council of Education of Castile and Leon under project grant VA257P18, and by the European Commission under project grant 588438-EPP-1-2017-1-EL-EPPKA2-KA. The rst author is supported by the European Social Fund and the Regional Council of Education of Castile and Leon.