Periodic Reporting for period 2 - AFEL (AFEL - Analytics For Everyday Learning)
Berichtszeitraum: 2017-06-01 bis 2018-11-30
Research on measuring learners' online activities often focus on the restricted scenario of students formally engaged in learning and where online activities happen through a provided eLearning system. However, more and more learners are now using a large variety of online platforms and resources which are not necessarily connected with their learning environment or with each-other. Such use of online resources tends to be self-directed in the sense that learners make their own choices as to which resource to employ and which activity to carry out. With such practices becoming more common, the goal of AFEL (Analytics for Everyday Learning) was to address both the theoretical and technological challenges arising from applying learning analytics in the context of online, self-directed learning. The pillars of the project are the technologies to capture large scale, heterogeneous data about learners' online activities across multiple platforms (including social media) and the operationalization of theoretical cognitive models of learning to analyse those online learning activities. One of the key outcomes of the project is therefore a set of tools enabling learners to track online learning activities so to enable them to reflect and ultimately improve the way they either focus their own learning or help others to learn. This is achieved through confronting theoretical, cognitive models originating from learning psychology, collaboration and social interactions to the actual practices of online learning as captured by technological means. Indeed, in this way, AFEL developed a complete technology stack, from platforms for capturing data about learning activities on a large variety of online systems and resources, to advanced and personalized visualizations, to enable end-users engaged with learning to reflect on their learning behaviour, and on how to best achieve their learning goals. Those tools and technologies were developed to be transferable to a variety of online learning platform, and tested through the Didactalia platform as well as the more general, and more challenging, scenario of learning through open web browsing.
"There are three main aspects to the work in AFEL: 1- The underlying, theoretical cognitive model for online learning, 2- the technical framework for capturing, processing and (visually) analysing data related to online learning, and 3- the integration of those two aspects into a set of usable tools for online learners.
The AFEL project relies on the co-evolution model where learning and knowledge construction happen as a result of the interaction between the cognitive system of the learner and the social system in which they operate. On this basis, an entity model was created that encapsulates the interaction between different actors in online learning, supporting data capture. The co-evolution model also represents a basis to study progress in online learning, including, besides the classical view of learning as integrating new ""topical"" knowledge, aspects related to the complexity of the activities being performed or the variety of views they might encapsulate. Specific studies are therefore being conducted in order to investigate those aspects (including studies of the ""echo chamber"" effect in online communities). As a side effect of the theoretical work in the project being at the basis of the technical work, an internal ""glossary"" was created to include shared definitions of the common concepts being manipulated in the project. The model, glossary and learning from the project have not only driven the design of the AFEL tools and studies, they are also being taken forward in further research on informal learning through online resources.
A key challenge for the AFEL project was the need for a technological infrastructure capable of collecting, processing and analysing large amounts of data from distributed, heterogeneous sources. AFEL created a robust and highly transferable data platform: A software architecture to collect, store, process and make data available from various sources of data about learning activities and resources. This architecture implements a plugeable pipeline to collect large, high-velocity online activity data, enrich those data to support classification, analysis and recommendation, compute indicators of learning and recommendations, and package the data for delivery to various interfaces, including the AFEL mobile application and visual analytics dashboard. This platform has demonstrated its ability to support capturing everyday learning activities through evaluation studies in the context of the Didactalia platform, and in the more challenging context of open web browsing, where large amounts of online activity data was collected and process for a long period of time. This has allowed us to demonstrate its transferrability and to clearly scope its exploitation, as part of GNOSS services and products, as well as as an add-on to other online learning platforms.
The overall goal of the work in AFEL was the creation of a set of tools that can demonstrably help online learners in keeping track of their learning activities and be more effective in achieving their learning goals. Those tools have been built through multiple iterations, leading to a validated toolkit that has been shown to be usable in the context of learning through an online platform. Those include a mobile learning applications supporting users in monitoring their learning activities, a recommender engine that can be flexibly adapted to provide recommendations from various sources of learning material, and a visual analytics dashboard which provide and explorable view over the learner's data, which can be further customised by the learners themselves. As for the data platform on which they rely, those tools can be transferred to a variety of online learning platform, and customised/re-combined depending on the specific scenario and requirements of the learners/learning platform. They have been integrated in the product/service offering of GNOSS and will be exploited through deployments supporting organisations relying on other le"
The AFEL project relies on the co-evolution model where learning and knowledge construction happen as a result of the interaction between the cognitive system of the learner and the social system in which they operate. On this basis, an entity model was created that encapsulates the interaction between different actors in online learning, supporting data capture. The co-evolution model also represents a basis to study progress in online learning, including, besides the classical view of learning as integrating new ""topical"" knowledge, aspects related to the complexity of the activities being performed or the variety of views they might encapsulate. Specific studies are therefore being conducted in order to investigate those aspects (including studies of the ""echo chamber"" effect in online communities). As a side effect of the theoretical work in the project being at the basis of the technical work, an internal ""glossary"" was created to include shared definitions of the common concepts being manipulated in the project. The model, glossary and learning from the project have not only driven the design of the AFEL tools and studies, they are also being taken forward in further research on informal learning through online resources.
A key challenge for the AFEL project was the need for a technological infrastructure capable of collecting, processing and analysing large amounts of data from distributed, heterogeneous sources. AFEL created a robust and highly transferable data platform: A software architecture to collect, store, process and make data available from various sources of data about learning activities and resources. This architecture implements a plugeable pipeline to collect large, high-velocity online activity data, enrich those data to support classification, analysis and recommendation, compute indicators of learning and recommendations, and package the data for delivery to various interfaces, including the AFEL mobile application and visual analytics dashboard. This platform has demonstrated its ability to support capturing everyday learning activities through evaluation studies in the context of the Didactalia platform, and in the more challenging context of open web browsing, where large amounts of online activity data was collected and process for a long period of time. This has allowed us to demonstrate its transferrability and to clearly scope its exploitation, as part of GNOSS services and products, as well as as an add-on to other online learning platforms.
The overall goal of the work in AFEL was the creation of a set of tools that can demonstrably help online learners in keeping track of their learning activities and be more effective in achieving their learning goals. Those tools have been built through multiple iterations, leading to a validated toolkit that has been shown to be usable in the context of learning through an online platform. Those include a mobile learning applications supporting users in monitoring their learning activities, a recommender engine that can be flexibly adapted to provide recommendations from various sources of learning material, and a visual analytics dashboard which provide and explorable view over the learner's data, which can be further customised by the learners themselves. As for the data platform on which they rely, those tools can be transferred to a variety of online learning platform, and customised/re-combined depending on the specific scenario and requirements of the learners/learning platform. They have been integrated in the product/service offering of GNOSS and will be exploited through deployments supporting organisations relying on other le"
Establishing both the theoretical and technological bases to enable the analysis of online learning has led to a varied set of research results, translated into high quality publications in renowned scientific venues. The particular strand represented by AFEL in the learning analytics research community has generated much interest in this community, as attested by the invitations received by key members of the project. Beyond the academic community, much of the impact generated by AFEL is driven by the tools and applications developed as part of the project. Those tools have been developed to be easy to deploy and customize by specific organisations. They are to be adopted by customers engaged with the commercial activities of GNOSS, and by organisations relying on other online learning platforms for which they can be adapted. Besides the tools, the release of the AFEL learning analytics dataset, as well as the development of the AFEL glossary represent key resources for researchers and practitioners of learning analytics which are expected to bring further benefits from the project to the community at large. Those represent reusable material to enable further research in education science, learning psychology, technology enhanced learning and learning analytics.