A novel framework of training hidden Markov support vector machines from lightly-annotated data

Natural language understanding (NLU) aims to map sentences to their semantic mean representations. Statistical approaches to NLU normally require fully-annotated training data where each sentence is paired with its word-level semantic annotations. In this paper, we propose a novel learning framework which trains the Hidden Markov Support Vector Machines (HM-SVMs) without the use of expensive fully-annotated data. In particular, our learning approach takes as input a training set of sentences labeled with abstract semantic annotations encoding underlying embedded structural relations and automatically induces derivation rules that map sentences to their semantic meaning representations. The proposed approach has been tested on the DARPA Communicator Data and achieved 93.18% in F-measure, which outperforms the previously proposed approaches of training the hidden vector state model or conditional random fields from unaligned data, with a relative error reduction rate of 43.3% and 10.6% being achieved.

Semi-supervised learning of statistical models for natural language understanding

Natural language understanding is to specify a computational model that maps sentences to their semantic mean representation. In this paper, we propose a novel framework to train the statistical models without using expensive fully annotated data. In particular, the input of our framework is a set of sentences labeled with abstract semantic annotations. These annotations encode the underlying embedded semantic structural relations without explicit word/semantic tag alignment. The proposed framework can automatically induce derivation rules that map sentences to their semantic meaning representations. The learning framework is applied on two statistical models, the conditional random fields (CRFs) and the hidden Markov support vector machines (HM-SVMs). Our experimental results on the DARPA communicator data show that both CRFs and HM-SVMs outperform the baseline approach, previously proposed hidden vector state (HVS) model which is also trained on abstract semantic annotations. In addition, the proposed framework shows superior performance than two other baseline approaches, a hybrid framework combining HVS and HM-SVMs and discriminative training of HVS, with a relative error reduction rate of about 25% and 15% being achieved in F-measure.

Organisational change and safety culture: the impact of communication

This research examines and explains the links between safety culture and communication. Safety culture is a concept that in recent years has gained prominence but there has been little applied research conducted to investigate the meaning of the concept in 'real life' settings. This research focused on a Train Operating Company undergoing change in a move towards privatisation. These changes were evident in the management of safety, the organisation of the industry and internally in their management. The Train Operating Company's management took steps to improve their safety culture and communications through the development of a cascade communication structure. The research framework employed a qualitative methodology in order to investigate the effect of the new system on safety culture. Findings of the research were that communications in the organisation failed to be effective for a number of reasons, including both cultural and logistical problems. The cultural problems related to a lack of trust in the organisation by the management and the workforce, the perception of communications as management propaganda, and asyntonic communications between those involved, whilst logistical problems related to the inherent difficulties of communicating over a geographically distributed network. An organisational learning framework was used to explain the results. It is postulated that one of the principal reasons why change, either to the safety culture or to communications, did not occur was because of the organisation's inability to learn. The research has also shown the crucial importance of trust between the members of the organisation, as this was one of the fundamental reasons why the safety culture did not change, and why safety management systems were not fully implemented. This is consistent with the notion of mutual trust in the HSC (1993) definition of safety culture. This research has highlighted its relevance to safety culture and its importance for organisational change.

Jointly event extraction and visualization on Twitter via probabilistic modelling

Event extraction from texts aims to detect structured information such as what has happened, to whom, where and when. Event extraction and visualization are typically considered as two different tasks. In this paper, we propose a novel approach based on probabilistic modelling to jointly extract and visualize events from tweets where both tasks benefit from each other. We model each event as a joint distribution over named entities, a date, a location and event-related keywords. Moreover, both tweets and event instances are associated with coordinates in the visualization space. The manifold assumption that the intrinsic geometry of tweets is a low-rank, non-linear manifold within the high-dimensional space is incorporated into the learning framework using a regularization. Experimental results show that the proposed approach can effectively deal with both event extraction and visualization and performs remarkably better than both the state-of-the-art event extraction method and a pipeline approach for event extraction and visualization.

An exploratory framework for analysing international retail learning

It is argued that international retail research has overlooked an essential component of the retail internationalization process, notably learning. This paper proposes an exploratory framework that enables the application of learning theory to the study of international retailing. The paper provides a meaningful starting point for developing an overarching framework which would represent one sort of re-conceptualization of the retail internationalization process, and arguably a new perspective for reinterpreting, re-evaluating and refining the existing literature on international retailing. Alongside this exploratory framework, we present a series of research propositions that might serve as an agenda for research into international retail learning. The paper concludes with a summary of the key themes and ways in which the area of international retail learning may be investigated.

National system of innovation and technological learning:an integrated framework for understanding technological capability development in the Russian forestry sector

One of the issues in the innovation system literature is examination of technological learning strategies of laggard nations. Two distinct bodies of literature have contributed to our insight into forces driving learning and innovation, National Systems of Innovation (NSI) and technological learning literature. Although both literatures yield insights on catch-up strategies of 'latecomer' nations, the explanatory powers of each literature by itself is limited. In this paper, a possible way of linking the macro- and the micro-level approaches by incorporating enterprises as active learning entities into the learning and innovation system is proposed. The proposed model has been used to develop research hypotheses and indicate research directions and is relevant for investigating the learning strategies of firms in less technologically intensive industries outside East Asia.

Engaging employers as partners in work-based learning assessment:proposal for a quality enhancement framework

This paper seeks to advance research and practice related to the role of employers in all stages of the assessment process of work-based learning (WBL) within a tripartite relationship of higher education institution (HEI), student and employer. It proposes a research-informed quality enhancement framework to develop good practice in engaging employers as partners in assessment. The Enhancement Framework comprises three dimensions, each of which includes elements and questions generated by the experiences of WBL students, HEI staff and employers. The three dimensions of the Enhancement Framework are: 1. ‘premises of assessment’ encompassing issues of learning, inclusion, standards and value; 2. ‘practice’, encompassing stages of assessment made up of course design, assessment task, responsibilities, support, grading and feedback; 3. ‘communication of assessment’ with the emphasis on role clarity, language and pathways. With its prompt questions, the Enhancement Framework may be used as a capacity-building tool for promoting, sustaining, benchmarking and evaluating productive dialogue and critical reflection about assessment between WBL partners. The paper concludes by emphasising the need for professional development as well as policy and research development, so that assessment in WBL can more closely correspond to the potentially transformative nature of the learning experience.

Developing a robust self evaluation framework for active learning:the first stage of an ERASMUS+ project (QAEMarketPlace4HEI)

In ensuring the quality of learning and teaching in Higher Education, self-evaluation is an important component of the process. An example would be the approach taken within the CDIO community whereby self-evaluation against the CDIO standards is part of the quality assurance process. Eight European universities (Reykjavik University, Iceland; Turku University of Applied Sciences, Finland; Aarhus University, Denmark; Helsinki Metropolia University of Applied Sciences, Finland; Ume? University, Sweden; Telecom Bretagne, France; Aston University, United Kingdom; Queens University Belfast, United Kingdom) are engaged in an EU funded Erasmus + project that is exploring the quality assurance process associated with active learning. The development of a new self-evaluation framework that feeds into a ?Marketplace? where participating institutions can be paired up and then engage in peer evaluations and sharing around each institutions approach to and implementation of active learning. All of the partner institutions are engaged in the application of CDIO within their engineering programmes and this has provided a common starting point for the partnership to form and the project to be developed. Although the initial focus will be CDIO, the longer term aim is that the approach could be of value beyond CDIO and within other disciplines. The focus of this paper is the process by which the self-evaluation framework is being developed and the form of the draft framework. In today?s Higher Education environment, the need to comply with Quality Assurance standards is an ever present feature of programme development and review. When engaging in a project that spans several countries, the wealth of applicable standards and guidelines is significant. In working towards the development of a robust Self Evaluation Framework for this project, the project team decided to take a wide view of the available resources to ensure a full consideration of different requirements and practices. The approach to developing the framework considered: a) institutional standards and processes b) national standards and processes e.g. QAA in the UK c) documents relating to regional / global accreditation schemes e.g. ABET d) requirements / guidelines relating to particular learning and teaching frameworks e.g. CDIO. The resulting draft self-evaluation framework is to be implemented within the project team to start with to support the initial ?Marketplace? pairing process. Following this initial work, changes will be considered before a final version is made available as part of the project outputs. Particular consideration has been paid to the extent of the framework, as a key objective of the project is to ensure that the approach to quality assurance has impact but is not overly demanding in terms of time or paperwork. In other words that it is focused on action and value added to staff, students and the programmes being considered.

Entrepreneurship and innovation through experience:student and employer perspectives of the value of work-based learning in promoting entrepreneurialism

The 'internationalisation' of Business and Management education, reflective of EU enlargement and the unprecedented globalisation of education, has resulted in growing numbers of overseas students adding a diversity and richness to the learning environment within many contemporary European Higher Educational Institutions (Green, 2006, Sliwa & Grandy, 2006). However, cross-national studies analyzing the impact that the internationalisation of business education has on the employability of business and management graduates are rare. Furthermore, there exists a notable gap in research aimed at identifying and conceptualising the generic business skills and competencies required by European employers of business and management graduates. By proposing a conceptual framework based upon a working model of business graduate employability, this goes some way to addressing this gap.

Adaptive back-propagation in on-line learning of multilayer networks

An adaptive back-propagation algorithm is studied and compared with gradient descent (standard back-propagation) for on-line learning in two-layer neural networks with an arbitrary number of hidden units. Within a statistical mechanics framework, both numerical studies and a rigorous analysis show that the adaptive back-propagation method results in faster training by breaking the symmetry between hidden units more efficiently and by providing faster convergence to optimal generalization than gradient descent.

The learning dynamics of a universal approximator

The learning properties of a universal approximator, a normalized committee machine with adjustable biases, are studied for on-line back-propagation learning. Within a statistical mechanics framework, numerical studies show that this model has features which do not exist in previously studied two-layer network models without adjustable biases, e.g., attractive suboptimal symmetric phases even for realizable cases and noiseless data.

The exact sample complexity of PAC-learning problems with unit VC dimension

The Vapnik-Chervonenkis (VC) dimension is a combinatorial measure of a certain class of machine learning problems, which may be used to obtain upper and lower bounds on the number of training examples needed to learn to prescribed levels of accuracy. Most of the known bounds apply to the Probably Approximately Correct (PAC) framework, which is the framework within which we work in this paper. For a learning problem with some known VC dimension, much is known about the order of growth of the sample-size requirement of the problem, as a function of the PAC parameters. The exact value of sample-size requirement is however less well-known, and depends heavily on the particular learning algorithm being used. This is a major obstacle to the practical application of the VC dimension. Hence it is important to know exactly how the sample-size requirement depends on VC dimension, and with that in mind, we describe a general algorithm for learning problems having VC dimension 1. Its sample-size requirement is minimal (as a function of the PAC parameters), and turns out to be the same for all non-trivial learning problems having VC dimension 1. While the method used cannot be naively generalised to higher VC dimension, it suggests that optimal algorithm-dependent bounds may improve substantially on current upper bounds.

Globally optimal parameters for on-line learning in multilayer neural networks

We present a framework for calculating globally optimal parameters, within a given time frame, for on-line learning in multilayer neural networks. We demonstrate the capability of this method by computing optimal learning rates in typical learning scenarios. A similar treatment allows one to determine the relevance of related training algorithms based on modifications to the basic gradient descent rule as well as to compare different training methods.

Globally optimal on-line learning rules for multi-layer neural networks

We present a method for determining the globally optimal on-line learning rule for a soft committee machine under a statistical mechanics framework. This rule maximizes the total reduction in generalization error over the whole learning process. A simple example demonstrates that the locally optimal rule, which maximizes the rate of decrease in generalization error, may perform poorly in comparison.

The role of biases in on-line learning of two-layer networks

The influence of biases on the learning dynamics of a two-layer neural network, a normalized soft-committee machine, is studied for on-line gradient descent learning. Within a statistical mechanics framework, numerical studies show that the inclusion of adjustable biases dramatically alters the learning dynamics found previously. The symmetric phase which has often been predominant in the original model all but disappears for a non-degenerate bias task. The extended model furthermore exhibits a much richer dynamical behavior, e.g. attractive suboptimal symmetric phases even for realizable cases and noiseless data.