In this thesis, I present, illustrate and empirically validate a novel approach to modelling and explaining design process. The main outcome of this work is the formal definition of the problem framing, and the formulation of a recursive model of framing in design. The model (code-named RFD), represents a formalisation of a grey area in the science of design, and sees the design process as a recursive interaction of problem framing and problem solving. The proposed approach is based upon a phenomenon introduced in cognitive science and known as (reflective) solution talkback. Previously, there were no formalisations of the knowledge interactions occurring within this complex reasoning operation. The recursive model is thus an attempt to express the existing knowledge in a formal and structured manner. In spite of rather abstract, knowledge level on which the model is defined, it is a firm step in the clarification of design process. The RFD model is applied to the knowledge-level description of the conducted experimental study that is annotated and analysed in the defined terminology. Eventually, several schemas implied by the model are identified, exemplified, and elaborated to reflect the empirical results. The model features the mutual interaction of predicates ‘specifies’ and ‘satisfies’. The first asserts that a certain set of explicit statements is sufficient for expressing relevant desired states the design is aiming to achieve. The validity of predicate ‘specifies’ might not be provable directly in any problem solving theory. A particular specification can be upheld or rejected only by drawing upon the validity of a complementary predicate ‘satisfies’ and the (un-)acceptability of the considered candidate solution (e.g. technological artefact, product). It is the role of the predicate ‘satisfies’ to find and derive such a candidate solution. The predicates ‘specifies’ and ‘satisfies’ are contextually bound and can be evaluated only within a particular conceptual frame. Thus, a solution to the design problem is sound and admissible with respect to an explicit commitment to a particular specification and design frame. The role of the predicate ‘acceptable’ is to compare the admissible solutions and frames against the ‘real’ design problem. As if it answered the question: “Is this solution really what I wanted/intended?” Furthermore, I propose a set of principled schemas on the conceptual (knowledge) level with an aim to make the interactive patterns of the design process explicit. These conceptual schemas are elicited from the rigorous experiments that utilised the structured and principled approach to recording the designer’s conceptual reasoning steps and decisions. They include • the refinement of an explicit problem specification within a conceptual frame; • the refinement of an explicit problem specification using a re-framed reference; and • the conceptual re-framing (i.e. the identification and articulation of new conceptual terms) Since the conceptual schemas reflect the sequence of the ‘typical’ decisions the designer may make during the design process, there is no single, symbol-level method for the implementation of these conceptual patterns. Thus, when one decides to follow the abstract patterns and schemas, this abstract model alone can foster a principled design on the knowledge level. It must be acknowledged that for the purpose of computer-based support, these abstract schemas need to be turned into operational models and consequently suitable methods. However, such operational perspective was beyond the time and resource constraints placed on this research.

Research in patient education suggests that tailored educational materials can improve patient's understanding of a treatment plan and help to achieve patient engagement and compliance. The goal of the HealthDoc Project has been the creation of automated Natural Language Generation systems for producing educational materials that are tailored to an individual patient's medical condition and personal situation. The project has so far focused on developing computational linguistic tools needed to author tailorable content from which customized versions could be generated. Also the HealthDoc model of document generation assumes the existence of previously authored textual material. Therefore, a new approach is needed to construct these materials and ensure that the relevant medical knowledge will be captured and delivered to the patient by providing a means to assist the health care professionals in directly authoring the required domain knowledge. <br /><br />We have used constructivist educational theory and knowledge-level modelling to define a new approach incorporating Patient-centric and Behaviour-modifying Educational Model (PBEM) and a knowledge-acquisition framework. Unlike traditional approaches, in which all patients are treated alike in terms of the medical information provided, our new model takes into account characteristics of individual patients. This facilitates the patient's assimilation of relevant information pertaining to her behaviour and health. As the information provided must address the various concerns of different stakeholders, and different patients have different concerns and concern intensities, a knowledge-acquisition framework was developed to provide a structure for patient knowledge acquisition. This framework includes the following components: a Strategic Model, a Concerns Model, and an Interrogation-based knowledge-acquisition Tool. The tool is intended to be used directly by health care professionals and to assist them in formulating, structuring, representing, and articulating their domain knowledge. This research work explores a new field, knowledge-level modelling, for generating patient-tailored educational materials and provides guidelines to implementing such a knowledge-acquisition tool.

Research in patient education suggests that tailored educational materials can improve patient's understanding of a treatment plan and help to achieve patient engagement and compliance. The goal of the HealthDoc Project has been the creation of automated Natural Language Generation systems for producing educational materials that are tailored to an individual patient's medical condition and personal situation. The project has so far focused on developing computational linguistic tools needed to author tailorable content from which customized versions could be generated. Also the HealthDoc model of document generation assumes the existence of previously authored textual material. Therefore, a new approach is needed to construct these materials and ensure that the relevant medical knowledge will be captured and delivered to the patient by providing a means to assist the health care professionals in directly authoring the required domain knowledge. <br /><br />We have used constructivist educational theory and knowledge-level modelling to define a new approach incorporating Patient-centric and Behaviour-modifying Educational Model (PBEM) and a knowledge-acquisition framework. Unlike traditional approaches, in which all patients are treated alike in terms of the medical information provided, our new model takes into account characteristics of individual patients. This facilitates the patient's assimilation of relevant information pertaining to her behaviour and health. As the information provided must address the various concerns of different stakeholders, and different patients have different concerns and concern intensities, a knowledge-acquisition framework was developed to provide a structure for patient knowledge acquisition. This framework includes the following components: a Strategic Model, a Concerns Model, and an Interrogation-based knowledge-acquisition Tool. The tool is intended to be used directly by health care professionals and to assist them in formulating, structuring, representing, and articulating their domain knowledge. This research work explores a new field, knowledge-level modelling, for generating patient-tailored educational materials and provides guidelines to implementing such a knowledge-acquisition tool.