Search

Assessment quality and validity is heavily reliant on the quality of items included in an assessment or test. Difficulty is an essential factor that can determine items and tests’ overall quality. Therefore, item difficulty prediction is extremely important in any pedagogical learning environment. Data-driven approaches to item difficulty prediction are gaining more and more prominence, as demonstrated by the recent literature. In this paper, we provide a systematic review of data-driven approaches to item difficulty prediction. Of the 148 papers that were identified that cover item difficulty prediction, 38 papers were selected for the final analysis. A classification of the different approaches used to predict item difficulty is presented, together with the current practices for item difficulty prediction with respect to the learning algorithms used, and the most influential difficulty features that were investigated.

Type 2 diabetes is one of the most serious chronic diseases which can be managed by medication and lifestyle changes. Regular physical activity is an example of lifestyle modification that can help in managing and preventing complications of the disease. However, a number of barriers to physical activity of different origin and type (e.g. health, personal, and psychological barriers) can prevent patients from achieving their goals. Various studies have attempted to categories the different barriers, but there is no unified model representing the different barriers and the possible interactions between them and the patient's activities. In this paper, we propose a conceptual model to identify and classify the barriers to physical activity for type 2 diabetes that is intended lay the foundations for the development of an ontology, i.e. a formal model of barriers and their relationships with diseases and patient's activities. The proposed model relies on identifying and classifying the barriers to physical activity according to their signs or factors, and reuses existing formal models of diabetes and other open source specialised resources.

Digital behavior intervention is a growing area of research which investigates how interactive systems can encourage and support people to change their behavior. Personalization plays an important role in this, as the most effective persuasive and motivational strategies are likely to depend on user characteristics such as the user’s personality, affective state, existing attitudes, behaviors, knowledge, and goals. Example application areas include healthcare (e.g., encouraging people to eatmore healthily and exercise more), education (e.g., motivating learners to study more), environment (e.g., encouraging people to use less energy and more public transport), and collaborative content development (e.g., incentivizing people to annotate resources). The idea that a system may be used to motivate people on the basis of a user model is certainly not novel. As early as the 1980s, intelligent tutoring systems encouraged students to learn by means of tailored feedback and hints (Sleeman and Brown 1982), and in the 90s patient education systems attempted to address compliance to a medical regimen by means of information and personalized advice (De Carolis et al. 1996) or encouraged people to adopt healthier lifestyles (Reiter et al. 2003). However, it is only recently that a number of seemingly non correlated, extensive research efforts

Whilst significant attention has been given to centralised approaches for aligning full ontologies, limited attention has been given to the problem of aligning partially exposed ontologies in a decentralised setting. Traditional ontology alignment techniques rely on the full disclosure of the ontological models that find the "best" set of correspondences that map entities from one ontology to another. However, within open and opportunistic environments, such approaches may not always be pragmatic or even acceptable due to privacy concerns. We present a novel dialogue based negotiation mechanism that supports the strategic agreement over correspondences between agents with limited or no prior knowledge of their opponent's ontology. This mechanism allows both agents to reach a mutual agreement over an alignment through the selective disclosure of their ontological model, and facilitates rational choices on the grounds of their ontological knowledge and their specific strategies. We formally introduce the dialogue mechanism, and discuss its behaviour, properties and outcomes.

Conflict situations do not only arise from misunderstandings, erroneous perceptions, partial knowledge, false beliefs, etc., but also from differences in “opinions” and in the different agents' value systems. It is not always possible, and maybe not even desirable, to “solve” this kind of conflict, as the sources are subjective. The communicating agents can, however, use knowledge of the opponent's preferences, to try and convince the partner of a point of view which they wish to promote. To deal with these situations requires an argumentative capacity, able to handle not only “demonstrative” arguments but also “dialectic” ones, which may not necessarily be based on rationality and valid premises. This paper presents a formalization of a theory of informal argumentation, focused on techniques to change attitudes of the interlocutor, in the domain of health promotion.

In this paper, we describe how user-adapted explanations about drug prescriptions can be generated from already existing data sources. We start by illustrating the two-step approach employed in the first version of the natural language generator and the limitations of generated texts, that we discovered through analytical and empirical evaluations. We claim that, although style refinement would be needed in these texts, particular care should be devoted to implementing some of the persuasion techniques that doctors employ in their explanations. This would require either thoroughly revising the text planning techniques employed or converting to a multistep generation architecture. We justify why we selected this second alternative and propose some heuristics to repair problems found in the first version of the generator. Some final considerations about the advantages of this approach and the possibility of generalizing it to other domains conclude the paper.

This paper describes a computer assisted interview for children and vulnerable adults. The system implements a “triadic interview” interaction since it is used as a focus point between the child and the clinician, whose dialogue is mediated by the tool. The tool helps children express their feelings and experiences, by making use of an “emotion palette” and a set of sub-tools developed on paper by children and transformed into computer based depictions. The tool has been extensively evaluated in clinical practice, providing a strong indication of its ability to improve the quality of the interaction with children.

In this article, we examine the case of a system that cooperates with a “direct” user to plan an activity that some “indirect” user, not interacting with the system, should perform. The specific application we consider is the prescription of drugs. In this case, the direct user is the prescriber and the indirect user is the person who is responsible for performing the therapy. Relevant characteristics of the two users are represented in two user models. Explanation strategies are represented in planning operators whose preconditions encode the cognitive state of the indirect user; this allows tailoring the message to the indirect user's characteristics. Expansion of optional subgoals and selection among candidate operators is made by applying decision criteria represented as metarules, that negotiate between direct and indirect users' views also taking into account the context where explanation is provided. After the message has been generated, the direct user may ask to add or remove some items, or change the message style. The system defends the indirect user's needs as far as possible by mentioning the rationale behind the generated message. If needed, the plan is repaired and the direct user model is revised accordingly, so that the system learns progressively to generate messages suited to the preferences of people with whom it interacts.

The idea that a computer system could be used to motivate people to perform a certain task on the basis of a user model is certainly not novel. As early as the 80s, intelligent tutoring systems would encourage students to learn by means of tailored feedback and hints [24], and in the 90s patient education systems would attempt to address the problem of compliance to a medical regimen by means of information and personalised advice [1] or would encourage people to adopt healthier lifestyles [19]. It is however only recently that a number of, seemingly non correlated, extensive research efforts, from various perspectives, have started to focus on a more complex cognitive model of rational and extra-rational features, involving emotions, persuasion, motivation and argumentation. We can distinguish three parallel strands of research that have become prominent.

In this paper we present two different approaches to personality diagnosis, for the provision of innovative personalized services, as used in a case study where diabetic patients were supported in the improvement of physical activity in their daily life. The first approach presented relies on a static clustering of the population, with a specific motivation strategy designed for each cluster. The second approach relies on a dynamic population clustering, making use of recommendation systems and algorithms, like Collaborative Filtering. We discuss pro and cons of each approach and a possible combination of the two, as the most promising solution for this and other personalization services in eHealth.

The aim of this chapter is to explore the problems and challenges of creating a ‘Persuasion Machine’ that is intended to engage a user in an argument in order to persuade her of some point of view. The approach is to start with a focus upon breadth, and to bring structure to an ill-structured problem. From there, for each component and each subproblem, the linguistic, computational, rhetorical and argumentation-theoretic theory drivers are integrated, to develop scaffolding and solutions that work not only as an environment in which to assess new theoretical developments, but also as a route to implementation and evaluation.

Health promotion is a complex activity that requires both explanation and persuasion skills. This paper proposes a three-layered model of dialogue coding, based on a rhetorical argumentation model, and a behavioural model of change. The model was applied to the analysis of a corpus of 40 e-mail dialogue exchanges on healthy nutrition advice. Examples of analysis are given.

The notions of argument and argumentation have become increasingly ubiquitous in Artificial Intelligence research, with various application and interpretations. Less attention has been, however, specifically devoted to rhetorical argument The work presented in this paper aims at bridging this gap, by proposing a framework for characterising rhetorical argumentation, based on Perelman and Olbrechts-Tyteca's New Rhetoric. The paper provides an overview of the state of the art of computational work based on, or dealing with, rhetorical aspects of argumentation, before presenting the characterisation proposed, corroborated by walked-through examples.

This paper offers an introduction to the 2001 Workshop on Computational Models of Natural Language Argument (CMNLA 2001), a special event of the International Conference on Computational Science. The contributors to the workshop represent, in their backgrounds, the diversity of fields upon which the focus of the event draws. As a result, this paper aims not only to introduce the accepted papers, but also to provide a background that will be accessible to researchers in the various fields, and to sit each work into a coherent context.

Some years ago, we started a research study devoted at investigating the reasoning processes involved in the dialogues between agents with personality; we focused, in particular, on conflict resolution dialogues. In choosing an application domain, we decided to exploit the medical literature, both for ‘historical’ reasons (our personal research interests) and because dialogues in this scenario are complex enough to provide interesting ideas without being trivial. In a previous work (de Rosis and Grasso, 1997; de Rosis et al., 1998b), we examined the styles of reasoning that guides such a dialogue and we formalized these styles in a dialogue simulation system, XANTHIPPE. The system, developed in Lisp, had a believable behaviour in most of the simpler situations. However, it was not able to deal with more sophisticated aspects of the reasoning process, such as elusion, reticence, omission and other stronger forms of deception.

Explanations are an important by-product of medical decisionsupport activities, as they have proved to favour compliance and correct treatment performance. To achieve this purpose, these texts should have a strong argumentation content and should adapt to emotional, as well as to rational attitudes of the Addressee. This paper describes how Rhetorical Sentence Planning can contribute to this aim: the rulebased plan discourse revision is introduced between Text Planning and Linguistic Realization, and exploits knowledge about the user personality and emotions and about the potential impact of domain items on user compliance and memory recall. The proposed approach originates from analytical and empirical evaluation studies of computer generated explanation texts in the domain of drug prescription.