Parameter learning for POMDP spoken dialogue models

The partially observable Markov decision process (POMDP) provides a popular framework for modelling spoken dialogue. This paper describes how the expectation propagation algorithm (EP) can be used to learn the parameters of the POMDP user model. Various special probability factors applicable to this task are presented, which allow the parameters be to learned when the structure of the dialogue is complex. No annotations, neither the true dialogue state nor the true semantics of user utterances, are required. Parameters optimised using the proposed techniques are shown to improve the performance of both offline transcription experiments as well as simulated dialogue management performance. ©2010 IEEE.

An investigation into inductive parameter learning in complex hierarchical knowledge structures representing clinical expertise

This dissertation investigates the very important and current problem of modelling human expertise. This is an apparent issue in any computer system emulating human decision making. It is prominent in Clinical Decision Support Systems (CDSS) due to the complexity of the induction process and the vast number of parameters in most cases. Other issues such as human error and missing or incomplete data present further challenges. In this thesis, the Galatean Risk Screening Tool (GRiST) is used as an example of modelling clinical expertise and parameter elicitation. The tool is a mental health clinical record management system with a top layer of decision support capabilities. It is currently being deployed by several NHS mental health trusts across the UK. The aim of the research is to investigate the problem of parameter elicitation by inducing them from real clinical data rather than from the human experts who provided the decision model. The induced parameters provide an insight into both the data relationships and how experts make decisions themselves. The outcomes help further understand human decision making and, in particular, help GRiST provide more accurate emulations of risk judgements. Although the algorithms and methods presented in this dissertation are applied to GRiST, they can be adopted for other human knowledge engineering domains.

Expression and connection : the integration of the reflective learning process and the writing process into social network sites

A number of instructors have recently adopted social network sites (SNSs) for learning. However, the learning design of SNSs often remains at a preliminary level similar to a personal log book because it does not properly include reflective learning elements such as individual reflection and collaboration. This article looks at the reflective learning process and the public writing process as a way of improving the quality of reflective learning on SNSs. It proposes a reflective learning model on SNSs based on two key pedagogical concepts for social networking: individual expression and collaborative connection. It is expected that the model would be helpful for instructors in designing a reflective learning process on SNSs in an effective and flexible way.

Building sustainable and effective assurance of learning process in a changing higher education environment

In a study of assuring learning in Australian Business Schools, 25 Teaching and Learning Associate Deans were interviewed to identify current issues in developing and measuring the quality of teaching and learning outcomes. Results indicate that for most institutions developing a perspective on graduate attributes and mapping assessments to measure outcomes across an entire program required knowledge creation and the building of new inclusive processes. Common elements of effective practice, namely those which offered consistently superior outcomes, included: inclusive processes; embedded graduate attributes throughout a program; alongside consistent and appropriate assessment. Results indicate that assurance of learning processes are proliferating nationally while quality of teaching and learning outcomes and in the processes for assuring it is increasing as a result.

A single-rate context-dependent learning process underlies rapid adaptation to familiar object dynamics.

Motor learning has been extensively studied using dynamic (force-field) perturbations. These induce movement errors that result in adaptive changes to the motor commands. Several state-space models have been developed to explain how trial-by-trial errors drive the progressive adaptation observed in such studies. These models have been applied to adaptation involving novel dynamics, which typically occurs over tens to hundreds of trials, and which appears to be mediated by a dual-rate adaptation process. In contrast, when manipulating objects with familiar dynamics, subjects adapt rapidly within a few trials. Here, we apply state-space models to familiar dynamics, asking whether adaptation is mediated by a single-rate or dual-rate process. Previously, we reported a task in which subjects rotate an object with known dynamics. By presenting the object at different visual orientations, adaptation was shown to be context-specific, with limited generalization to novel orientations. Here we show that a multiple-context state-space model, with a generalization function tuned to visual object orientation, can reproduce the time-course of adaptation and de-adaptation as well as the observed context-dependent behavior. In contrast to the dual-rate process associated with novel dynamics, we show that a single-rate process mediates adaptation to familiar object dynamics. The model predicts that during exposure to the object across multiple orientations, there will be a degree of independence for adaptation and de-adaptation within each context, and that the states associated with all contexts will slowly de-adapt during exposure in one particular context. We confirm these predictions in two new experiments. Results of the current study thus highlight similarities and differences in the processes engaged during exposure to novel versus familiar dynamics. In both cases, adaptation is mediated by multiple context-specific representations. In the case of familiar object dynamics, however, the representations can be engaged based on visual context, and are updated by a single-rate process.

Mapping experience in organizations: A learning process for strategic technology planning

Today's fast-paced, dynamic environments mean that for organizations to keep "ahead of the game", engineering managers need to maximize current opportunities and avoid repeating past mistakes. This article describes the development study of a collaborative strategic management tool - the Experience Scan to capture past experience and apply learning from this to present and future situations. Experience Scan workshops were held in a number of different technology organizations, developing and refining the tool until its format stabilized. From participants' feedback, the workshop-based tool was judged to be a useful and efficient mechanism for communication and knowledge management, contributing to organizational learning.

序破急 Jo-Ha-Kyu: enticement - crux - consolidation. From study to learning: Process drama projects in the Japanese English language university classroom

This dissertation investigates how social issues can be explored through process drama projects in the Japanese university English as a Foreign Language classroom context. The trajectory of this dissertation moves along a traditional Noh three part macro-continuum, called Jo-Ha-Kyu, interpreted as enticement, crux and consolidation. Within these three parts, there are six further divisions. Part I consists of three sections: Section I, the introduction, sets the backdrop for the entire dissertation, that of Japan, and aims to draw the reader into its culturally unique and specific world. This section outlines the rationale for placing the ethnographer at the centre of the research, and presents Japan through the eyes of the writer. Section II outlines relevant Japanese cultural norms, mores and values, the English educational landscape of Japan and an overview of theatre in Japan and its possible influences on the Japanese university student today. Section III provides three literature reviews: second language acquisition, drama in education to process drama, and Content Language Integrated Learning. In Part 2, Sections IV and V respectively consist of the research methodology and the action research at the core of this dissertation. Section IV describes the case of Kwansei Gakuin University, then explains the design of the process drama curricula. Section V details the three-process drama projects based around the three social issues at the centre of this dissertation. There is also a description of an extra project that of the guest lecturer project. The ultimate goals of all four projects were to change motivation through English in a CLIL context, to develop linguistic spontaneity and to deepen emotional engagement with the themes. Part 3 serves to reflect upon the viability of using process drama in the Japanese university curriculum, and to critically self-reflect on the project as a whole.

Improving parameter learning of Bayesian nets from incomplete data

This paper addresses the estimation of parameters of a Bayesian network from incomplete data. The task is usually tackled by running the Expectation-Maximization (EM) algorithm several times in order to obtain a high log-likelihood estimate. We argue that choosing the maximum log-likelihood estimate (as well as the maximum penalized log-likelihood and the maximum a posteriori estimate) has severe drawbacks, being affected both by overfitting and model uncertainty. Two ideas are discussed to overcome these issues: a maximum entropy approach and a Bayesian model averaging approach. Both ideas can be easily applied on top of EM, while the entropy idea can be also implemented in a more sophisticated way, through a dedicated non-linear solver. A vast set of experiments shows that these ideas produce significantly better estimates and inferences than the traditional and widely used maximum (penalized) log-likelihood and maximum a posteriori estimates. In particular, if EM is adopted as optimization engine, the model averaging approach is the best performing one; its performance is matched by the entropy approach when implemented using the non-linear solver. The results suggest that the applicability of these ideas is immediate (they are easy to implement and to integrate in currently available inference engines) and that they constitute a better way to learn Bayesian network parameters.

Promoting self-regulated learning in technology enhanced learning environments : keeping a digital track of the learning process

Tese de doutoramento (co-tutela), Psicologia (Psicologia da Educação), Faculdade de Psicologia da Universidade de Lisboa, Faculdade de Psicologia e de Ciências da Educação da Universidade de Coimbra, Technial University of Darmstadt, 2014

Unlocking the technology-mediated learning process : insights and articulating actions for an institutional approach

This paper is part of the results from the project "Implementation Strategies and Development of an Open and Distance Education System for the University of the Azores" funded by the European Social Fund. http://hdl.handle.net/10400.3/2327

Conceptualizing Institutional Capacity Building Through a Learning Process Set in Motion by a Foundation

Employing critical pedagogy and transformative theory as a theoretical framework, I examined a learning process associated with building capacity in community-based organizations (CBOs) through an investigation of the Institutional Capacity Building Program (ICBP) initiated by a Foundation. The study sought to: (a) examine the importance of institutional capacity building for individual and community development; (b) investigate elements of a process associated with a program and characteristics of a learning process for building capacity in CBOs; and (c) analyze the Foundation’s approach to synthesizing, systematizing, and sharing learning. The study used a narrative research design that included 3 one-on-one, hour-long interviews with 2 women having unique vantage points in ICBP: one is a program facilitator working at the Foundation and the other runs a CBO supported by the Foundation. The interviews’ semistructured questions allowed interviewees to share stories regarding their experience with the learning process of ICB and enabled themes to emerge from their day-to-day experience. Through the analysis of this learning process for institutional capacity building, a few lessons can be drawn from the experience of the Foundation.