Andreas Harrer

Integrated representations and small data: towards contextualized and embedded analytics tools for learners

We present an approach to support learners by means of visualization and contextualization of learning analytics interventions in the learning process. We follow up on conceptual work of colleagues and derive further design principles oriented towards learners as recipients of LA results. These are shown with implementations in two distinct projects to fulfill learners information in collaborative learning processes.

Planning for Life - Educate Students to Plan: Syntactic and Semantic Support of Planning Activities with a Visual Language

Organising their own learning processes is a hard challenge for students. This is especially pronounced in self-regulated, collaborative and constructionist learning situations where students are supposed to jointly create artefacts, plan their work and learn how to learn together. In this paper we present our work on the support of the planning activities. The support happens both at the syntactic level, i.e. how to make correct plans, and the semantic level, i.e. how to make meaningful plans. We describe our concept of a visual language to specify plans, our different support functionalities and demonstrate its use with an extended practical example as well as some empirical evidence and consequences for experimentation plans gained from practical classroom experimentation.

Applying Network Models and Network Analysis Techniques to the Study of Online Communities

“Network science” has evolved as a new overarching paradigm for studying the structure and evolution of networks of various natures providing a rich set of techniques for modeling and analysis. Related to online learning and knowledge-building communities, there is a specific interest in methods of social network analysis (SNA) including also the analysis of actor-artifact networks. The application of network models and network analysis techniques to online communities and collaboration in masses can serve various purposes such as the identification of central actors and roles, the detection and tracking of subcommunities, and the tracing of ideas in knowledge-building communities. This chapter focuses on describing the corresponding approaches in such a way as to demonstrate the basic ideas without actually going into mathematical details. The different approaches will be exemplified with recent applications to the study of networked collaboration, especially in learning and teaching contexts.

Using Real-Time Gaze Based Awareness Methods to Enhance Collaboration

Using eye-tracking in applications can be used to identify which areas are looked at by their users. In collaborative software this information can be transmitted to partners in real-time to provide an additional information channel. This paper compares different types of real-time gaze data visualizations. For this purpose, a study with three groups is conducted, who have to solve a collaborative puzzle. In every group the gaze data from each participant is recorded and visualized in a different way depending on the specific group condition. The aim is to evaluate a new context-based visualization to be able to make use of the known advantages of coordinate-based gaze data visualization outside of the domain of What-You-See-Is-What-I-See (WYSIWIS) interfaces.

A Design Proposal for Learner-Centered Visualisations of Learning Analytics in Collaborative Scenarios

Complex collaborative learning activities require the learners to self-regulate and reflect on their learning process. Learning Analytics can be a means to support them in these tasks. Yet, current approaches of representing analytics results are mainly oriented towards researchers or instructors of online courses. In this paper we will present our position on learner-centered visualisations for analytics and designs we implemented in the Metafora project to serve this purpose.

MeET US -- An Eyetracking Study to Evaluate Awareness Functionality in the CSCL System Metafora

In the Metafora project we designed a platform for collaborative learning and planning of learning challenges that make use of heterogeneous and multiple learning tools. These complex scenarios of up to 20 hours of learning require advanced awareness support functionality to allow effective collaboration. To evaluate the effectiveness of our awareness support features in Metafora we conducted an eye tracking study. In this study we applied synchronous eye tracking on dyads of learners in a controlled lab situation where awareness messages are sent to the dyad. We report about our results and discuss the method and its implications for future work.

Information Management for Adaptive Automotive Human Machine Interfaces

Adaptive Human Machine Interfaces (HMI) provide a substantial contribution to avoid information overload of the driver. This paper presents novel concepts for context-aware information management in the sector of adaptive automotive HMI. In particular, we describe an architectural design which supports an easy integration of multiple strategies to manipulate and to adapt the in-vehicle information flow depending on the current driving context. The presented architecture follows the blackboard design pattern and enables opportunistic reasoning in the automotive domain. We further present a novel fusion strategy for messages based on a taxonomic message model. The fusion strategy allows to combine two or more low-level messages and to replace them with a higher-level message in order to reduce the information load for the driver. The described concepts have been evaluated with the help of a driving simulator study with N = 41 test persons. The results show a significant reduction of the subjective workload if the information manager is applied.

Discovery of Structural and Temporal Patterns in MOOC Discussion Forums

This work aims to explore methods to investigate the structure of knowledge exchange in discussion forums in massive open online courses (MOOCs) explicitly taking into account changing patterns over time. The paper covers three different aspects of forum analysis combining different methods. First, an approach for the extraction of dynamic communication networks from forum data based on the classification of forum posts is presented that takes into account the information exchange relations between forum users. Second, measures that characterise users according to information seeking and information giving behaviour are introduced and the development of individual actors is analysed. Third, blockmodelling and tensor decomposition approaches for reducing a dynamic network to an interpretable macro-structure reflecting knowledge exchange between clusters of actors over time are evaluated. This allows for the analysis of the communication structure related to information exchange between participants of large scale online courses in different aspects. The utility of the analytics framework is demonstrated along two case studies on forum discussions in two MOOCs offered on the Coursera platform.

A Workbench for Visual Design of Executable and Re-usable Network Analysis Workflows

In this paper we introduce the concept of a web-based analytics workbench to support researchers of social networks in their analytic processes. Making explicit these processes allows for sound design, re-use, and automated execution using an authoring system for visual representations of these analytic workflows. The workbench is implemented according to a flexible technical framework in which external and newly-defined analytic components can be integrated and used in conjunction with other analytic components. As a showcase we discuss the integration of a complex analytic process using multi-relational blockmodeling on a well-known reference data set.

An integrated method for Adaptive automotive Human Machine Interfaces

Adaptive Human Machine Interfaces (HMI) provide a substantial contribution to avoid an information overload of the driver by adapting the information presentation in accordance to the current driving situation. Although a well-defined model serves as the base of any context-aware system, current approaches for adaptive HMIs take into account only a limited model of the driver, the vehicle and the environmental aspects. This paper presents a feasible approach to model, interpret and classify driving situations. We propose a taxonomy for modelling relevant situational information. An event-based interpretation mechanism allows a deduction of spatio-temporal relations. Based on the situation model, the current situation is identified by performing probabilistic inference. The proposed method serves as a base for a context-aware information management and is demonstrated with help of an example scenario.