Whereas most research in AI system explanation for healthcare applications looks at developing algorithmic explanations targeted at AI experts or medical professionals, the question we raise is: How do we build meaningful explanations for laypeople? And how does a meaningful explanation affect user’s trust perceptions? Our research investigates how the key factors affecting human-AI trust change in the light of human expertise, and how to design explanations specifically targeted at non-experts. By means of a stage-based design method, we map the ways laypeople understand AI explanations in a User Explanation Model. We also map both medical professionals and AI experts’ practice in an Expert Explanation Model. A Target Explanation Model is then proposed, which represents how experts’ practice and layperson’s understanding can be combined to design meaningful explanations. Design guidelines for meaningful AI explanations are proposed, and a prototype of AI system explanation for non-expert users in a breast cancer scenario is presented and assessed on how it affect users’ trust perceptions.

C ollaborative C omputer‐ S upported A rgument V isualization ( CCSAV ) has often been proposed as an alternative over more conventional, mainstream platforms for online discussion (e.g., online forums and wikis). CCSAV tools require users to contribute to the creation of a joint artifact (argument map) instead of contributing to a conversation. In this paper we assess empirically the effects of this fundamental design choice and show that the absence of conversational affordances and socially salient information in representation‐centric tools is detrimental to the users' collaboration experience. We report empirical findings from a study in which subjects using different collaborative platforms (a forum, an argumentation platform, and a socially augmented argumentation tool) were asked to discuss and predict the price of a commodity. By comparing users' experience across several metrics we found evidence that the collaborative performance decreases gradually when we remove conversational interaction and other types of socially salient information. We interpret these findings through theories developed in conversational analysis (common ground theory) and communities of practice and discuss design implications. In particular, we propose balancing the trade‐off between knowledge reification and participation in representation‐centric tools with the provision of social feedback and functionalities supporting meaning negotiation.

When assessing student essays, educators look for the students' ability to present and pursue well-reasoned and strong arguments. Such scholarly argumentation is often articulated by rhetorical metadiscourse. Educators will be necessarily examining metadiscourse in students' writing as signals of the intellectual moves that make their reasoning visible. Therefore students and educators could benefit from available powerful automated textual analysis that is able to detect rhetorical metadiscourse. However, there is a need to validate such technologies in higher education contexts, since they were originally developed in non-educational applications. This paper describes an evaluation study of a particular language analysis tool, the Xerox Incremental Parser (XIP), on undergraduate social science student essays, using the mark awarded as a measure of the quality of the writing. As part of this exploration, the study presented in this paper seeks to assess the quality of the XIP through correlational studies and multiple regression analysis.

The first international workshop on discourse-centric learning analytics (DCLA) took place at LAK13 in Leuven, Belgium. That workshop succeeded in its aim of catalysing ideas and building community connections between those working in this field of social learning analytics. It also proposed a mission statement for DCLA: to devise and validate analytics that look beyond surface measures in order to quantify linguistic proxies for deeper learning. This year, the focus of the second international DCLA workshop, like that of LAK14, is on the intersection of learning analytics research, theory and practice. Once researchers have developed and validated discourse-centric analytics, how can these be successfully deployed at scale to support learning?

This paper describes a novel analytics dashboard which visualises the key features of scholarly documents. The Dashboard aggregates the salient sentences of scholarly papers, their rhetorical types and the key concepts mentioned within these sentences. These features are extracted from papers through a Natural Language Processing (NLP) technology, called Xerox Incremental Parser (XIP). The XIP Dashboard is a set of visual analytics modules based on the XIP output. In this paper, we briefly introduce the XIP technology and demonstrate an example visualisation of the XIP Dashboard.

This workshop anticipates that an important class of learning analytic will emerge at the intersection of research into learning dynamics, online discussion platforms, and computational linguistics. Written discourse is arguably the primary class of data that can give us insights into deeper learning and higher order qualities such as critical thinking, argumentation, mastery of complex ideas, empathy, collaboration and interpersonal skills. Moreover, the ability to write in a scholarly manner is a core competence, often taking the form of discourse with oneself and the literature. Computational linguistics research has developed a rich array of tools for machine interpretation of human discourse, but work to develop these tools in the context of learning is at a relatively early stage. Moreover, there is a significant difference between designing tools to assist researchers in discourse analysis, and their deployment on platforms to provide meaningful analytics for the learners and educators who are conducting that discourse. This workshop aims to catalyse ideas and build community connections among those who want to shape this field.

Drawing on sociocultural discourse analysis and argumentation theory, we motivate a focus on learners' discourse as a promising site for identifying patterns of activity which correspond to meaningful learning and knowledge construction. However, software platforms must gain access to qualitative information about the rhetorical dimensions to discourse contributions to enable such analytics. This is difficult to extract from naturally occurring text, but the emergence of more-structured annotation and deliberation platforms for learning makes such information available. Using the Cohere web application as a research vehicle, we present examples of analytics at the level of individual learners and groups, showing conceptual and social network patterns, which we propose as indicators of meaningful learning.