Computational studies to understand the role of social learning in team familiarity and its effects on team performance

This paper concerns social learning modes and their effects on team performance. Social learning, such as by observing others' actions and their outcomes, allows members of a team to learn what other members know. Knowing what other members know can reduce task communication and co-ordination overhead, which helps the team to perform faster since members can devote their attention to their tasks. This paper describes agent-based simulation studies using a computational model that implements different social learning modes as parameters that can be controlled in the simulations. The results show that social learning from both direct and indirect observations positively contributes to learning about what others know, but the value of social learning is sensitive to prior familiarity such that minimum thresholds of team familiarity are needed to realise the benefits of social learning. This threshold increases with task complexity. These findings clarify the level of influence that sociality has on social learning and sets up a formal framework by which to conduct studies on how social context influences learning and group performance.

Nonnegative shared subspace learning and its application to social media retrieval

Although tagging has become increasingly popular in online image and video sharing systems, tags are known to be noisy, ambiguous, incomplete and subjective. These factors can seriously affect the precision of a social tag-based web retrieval system. Therefore improving the precision performance of these social tag-based web retrieval systems has become an increasingly important research topic. To this end, we propose a shared subspace learning framework to leverage a secondary source to improve retrieval performance from a primary dataset. This is achieved by learning a shared subspace between the two sources under a joint Nonnegative Matrix Factorization in which the level of subspace sharing can be explicitly controlled. We derive an efficient algorithm for learning the factorization, analyze its complexity, and provide proof of convergence. We validate the framework on image and video retrieval tasks in which tags from the LabelMe dataset are used to improve image retrieval performance from a Flickr dataset and video retrieval performance from a YouTube dataset. This has implications for how to exploit and transfer knowledge from readily available auxiliary tagging resources to improve another social web retrieval system. Our shared subspace learning framework is applicable to a range of problems where one needs to exploit the strengths existing among multiple and heterogeneous datasets.

Joint learning and dictionary construction for pattern recognition

We propose a joint representation and classification framework that achieves the dual goal of finding the most discriminative sparse overcomplete encoding and optimal classifier parameters. Formulating an optimization problem that combines the objective function of the classification with the representation error of both labeled and unlabeled data, constrained by sparsity, we propose an algorithm that alternates between solving for subsets of parameters, whilst preserving the sparsity. The method is then evaluated over two important classification problems in computer vision: object categorization of natural images using the Caltech 101 database and face recognition using the Extended Yale B face database. The results show that the proposed method is competitive against other recently proposed sparse overcomplete counterparts and considerably outperforms many recently proposed face recognition techniques when the number training samples is small.

Robust learning of discriminative projection for multicategory classification on the Stiefel manifold

Learning a robust projection with a small number of training samples is still a challenging problem in face recognition, especially when the unseen faces have extreme variation in pose, illumination, and facial expression. To address this problem, we propose a framework formulated under statistical learning theory that facilitates robust learning of a discriminative projection. Dimensionality reduction using the projection matrix is combined with a linear classifier in the regularized framework of lasso regression. The projection matrix in conjunction with the classifier parameters are then found by solving an optimization problem over the Stiefel manifold. The experimental results on standard face databases suggest that the proposed method outperforms some recent regularized techniques when the number of training samples is small.

Supervised subspace learning with multi-class Lagrangian SVM on the Grassmann manifold

Learning robust subspaces to maximize class discrimination is challenging, and most current works consider a weak connection between dimensionality reduction and classifier design. We propose an alternate framework wherein these two steps are combined in a joint formulation to exploit the direct connection between dimensionality reduction and classification. Specifically, we learn an optimal subspace on the Grassmann manifold jointly minimizing the classification error of an SVM classifier. We minimize the regularized empirical risk over both the hypothesis space of functions that underlies this new generalized multi-class Lagrangian SVM and the Grassmann manifold such that a linear projection is to be found. We propose an iterative algorithm to meet the dual goal of optimizing both the classifier and projection. Extensive numerical studies on challenging datasets show robust performance of the proposed scheme over other alternatives in contexts wherein limited training data is used, verifying the advantage of the joint formulation.

Learning discriminative sequence models from partially labelled data for activity recognition

Recognising daily activity patterns of people from low-level sensory data is an important problem. Traditional approaches typically rely on generative models such as the hidden Markov models and training on fully labelled data. While activity data can be readily acquired from pervasive sensors, e.g. in smart environments, providing manual labels to support fully supervised learning is often expensive. In this paper, we propose a new approach based on partially-supervised training of discriminative sequence models such as the conditional random field (CRF) and the maximum entropy Markov model (MEMM). We show that the approach can reduce labelling effort, and at the same time, provides us with the flexibility and accuracy of the discriminative framework. Our experimental results in the video surveillance domain illustrate that these models can perform better than their generative counterpart (i.e. the partially hidden Markov model), even when a substantial amount of labels are unavailable.

Human activity learning and segmentation using partially hidden discriminative models

Learning and understanding the typical patterns in the daily activities and routines of people from low-level sensory data is an important problem in many application domains such as building smart environments, or providing intelligent assistance. Traditional approaches to this problem typically rely on supervised learning and generative models such as the hidden Markov models and its extensions. While activity data can be readily acquired from pervasive sensors, e.g. in smart environments, providing manual labels to support supervised training is often extremely expensive. In this paper, we propose a new approach based on semi-supervised training of partially hidden discriminative models such as the conditional random field (CRF) and the maximum entropy Markov model (MEMM). We show that these models allow us to incorporate both labeled and unlabeled data for learning, and at the same time, provide us with the flexibility and accuracy of the discriminative framework. Our experimental results in the video surveillance domain illustrate that these models can perform better than their generative counterpart, the partially hidden Markov model, even when a substantial amount of labels are unavailable.

A matrix factorization framework for jointly analyzing multiple nonnegative data source

Nonnegative matrix factorization based methods provide one of the simplest and most effective approaches to text mining. However, their applicability is mainly limited to analyzing a single data source. In this paper, we propose a novel joint matrix factorization framework which can jointly analyze multiple data sources by exploiting their shared and individual structures. The proposed framework is flexible to handle any arbitrary sharing configurations encountered in real world data. We derive an efficient algorithm for learning the factorization and show that its convergence is theoretically guaranteed. We demonstrate the utility and effectiveness of the proposed framework in two real-world applications–improving social media retrieval using auxiliary sources and cross-social media retrieval. Representing each social media source using their textual tags, for both applications, we show that retrieval performance exceeds the existing state-of-the-art techniques. The proposed solution provides a generic framework and can be applicable to a wider context in data mining wherever one needs to exploit mutual and individual knowledge present across multiple data sources.

Zero-day malware detection based on supervised learning algorithms of API call signatures

Zero-day or unknown malware are created using code obfuscation techniques that can modify the parent code to produce offspring copies which have the same functionality but with different signatures. Current techniques reported in literature lack the capability of detecting zero-day malware with the required accuracy and efficiency. In this paper, we have proposed and evaluated a novel method of employing several data mining techniques to detect and classify zero-day malware with high levels of accuracy and efficiency based on the frequency of Windows API calls. This paper describes the methodology employed for the collection of large data sets to train the classifiers, and analyses the performance results of the various data mining algorithms adopted for the study using a fully automated tool developed in this research to conduct the various experimental investigations and evaluation. Through the performance results of these algorithms from our experimental analysis, we are able to evaluate and discuss the advantages of one data mining algorithm over the other for accurately detecting zero-day malware successfully. The data mining framework employed in this research learns through analysing the behavior of existing malicious and benign codes in large datasets. We have employed robust classifiers, namely Naïve Bayes (NB) Algorithm, k−Nearest Neighbor (kNN) Algorithm, Sequential Minimal Optimization (SMO) Algorithm with 4 differents kernels (SMO - Normalized PolyKernel, SMO – PolyKernel, SMO – Puk, and SMO- Radial Basis Function (RBF)), Backpropagation Neural Networks Algorithm, and J48 decision tree and have evaluated their performance. Overall, the automated data mining system implemented for this study has achieved high true positive (TP) rate of more than 98.5%, and low false positive (FP) rate of less than 0.025, which has not been achieved in literature so far. This is much higher than the required commercial acceptance level indicating that our novel technique is a major leap forward in detecting zero-day malware. This paper also offers future directions for researchers in exploring different aspects of obfuscations that are affecting the IT world today.

E-Simulations for educating the professions in blended learning environments

This chapter introduces digital, role-based simulations as an emerging and powerful educational approach for the professions and for broader workforce development purposes. It is acknowledged that simulations used for education, professional development, and training, have a long history of development and use. The focus is on digital simulations (e-simulations) situated in blended learning environments and the improved affordances of the newer digital media used via the web to enhance the value of their contribution to learning and teaching in professional and vocationally-oriented fields. This is an area which has received less attention in the whole “e-learning” literature compared with the voluminous body of knowledge and practice on computer-mediated communication, online community building, social networking, and various forms of online (usually automated) assessment. A framework of blended e-simulation design is outlined. The chapter concludes by examining what the future might hold for simulations in further and higher education, and ongoing work-based learning.

Improving outcomes of preschool language delay in the community : protocol for the Language for Learning randomised controlled trial

Background

Early language delay is a high-prevalence condition of concern to parents and professionals. It may result in lifelong deficits not only in language function, but also in social, emotional/behavioural, academic and economic well-being. Such delays can lead to considerable costs to the individual, the family and to society more widely. The Language for Learning trial tests a population-based intervention in 4 year olds with measured language delay, to determine (1) if it improves language and associated outcomes at ages 5 and 6 years and (2) its cost-effectiveness for families and the health care system.

Methods/Design

A large-scale randomised trial of a year-long intervention targeting preschoolers with language delay, nested within a well-documented, prospective, population-based cohort of 1464 children in Melbourne, Australia. All children received a 1.25-1.5 hour formal language assessment at their 4th birthday. The 200 children with expressive and/or receptive language scores more than 1.25 standard deviations below the mean were randomised into intervention or ‘usual care’ control arms. The 20-session intervention program comprises 18 one-hour home-based therapeutic sessions in three 6-week blocks, an outcome assessment, and a final feed-back/forward planning session. The therapy utilises a ‘step up-step down’ therapeutic approach depending on the child’s language profile, severity and progress, with standardised, manualised activities covering the four language development domains of: vocabulary and grammar; narrative skills; comprehension monitoring; and phonological awareness/pre-literacy skills. Blinded follow-up assessments at ages 5 and 6 years measure the primary outcome of receptive and expressive language, and secondary outcomes of vocabulary, narrative, and phonological skills.

Discussion

A key strength of this robust study is the implementation of a therapeutic framework that provides a standardised yet tailored approach for each child, with a focus on specific language domains known to be associated with later language and literacy. The trial responds to identified evidence gaps, has outcomes of direct relevance to families and the community, includes a well-developed economic analysis, and has the potential to improve long-term consequences of early language delay within a public health framework.

Language, experience and professional learning (what Walter Benjamin can teach us)

This essay raises questions about how language educators might construct and further develop their epistemology of practice in and through the situations in which they work from day to day. The occasion for this paper is our work as guest editors of a special issue of L-1: Educational Studies in Language and Literature, when we invited L1 teachers to reflect on the role that language plays in their professional learning, whether it be in the form of conversations with peers, reflective writing, or by other means. We begin this essay by locating our reflections within our current policy context, namely the standards-based reforms that have come to dominate educational thinking around the world, offering a brief critique of the values and attitudes embedded within them. We then outline a philosophical framework as an alternative to the world-view reflected by such reforms, focusing specifically on the work of Walter Benjamin. In the final sections, we review our work as guest editors of the special issue of L-1, reflecting on what we have learned from the papers we have assembled for this issue, and locating our learning within the philosophical framework that we have drawn from Benjamin. We argue that it is timely for language educators to articulate the assumptions that inhere within their work, in contradistinction to the common sense embedded in standards. Thus we might begin to reconceptualise the relation between language, experience and professional learning in opposition to the hegemony of standards.

A multidimensional framework for embedded academic skill development : transition pedagogy in social work

As a consequence of the widening participation agenda, student cohorts in Australian higher education are becoming increasingly diverse. While diversity is often characterised by a focus on culture or ethnicity, this variability also independently exists in regard to competence in academic skills (Dillon, 2007). Successfully developing discipline-specific academic skills is crucial to a student’s learning, progress and attainment in higher education. The growing recognition that students are entering Australian universities with varying levels of academic preparedness as a result of the widening participation agenda has made effective academic skill support even more important, since ‘access without a reasonable chance of success is an empty promise’ (International Associations of Universities, 2008, p. 1).

Activating teacher critical moments of learning through reflection

Mathematical modelling is one of the current focuses in the Singapore Mathematics Curriculum Framework. A multi-tiered teaching experiment using design research methodology was conducted to build teachers’ capacity in designing, facilitating, and evaluating learning during mathematical modelling tasks for Primary 5 students (aged 10-11). This paper illustrates the use of the retrospective analysis phase within design research cycles to activate a critical moment of teacher learning involving the interplay between questioning and listening during her first attempt at facilitating a mathematical modelling task. The teacher affirmed her deliberate focuses in the use of questions to (a) refine students’ models, (b) encourage articulation of student ideas in self-evaluation of the models, and (c) clarify and understand student reasoning. However, she also discovered the importance of interpretative listening in conjunction with questioning to promote more sophisticated mathematisation processes in model development. Implications from the use of the retrospective analysis phase on activating critical moments of learning during teacher education will be discussed.

Managing manufacturing technology investments: An intelligent learning system approach

This paper proposes an intelligent decision-support system for managing manufacturing technology investments. The intelligent system is a hybrid integration of two information processing modules: case-based reasoning and fuzzy ARTMAP – a supervised adaptive resonance theory (ART) neural network with a multi-dimensional map. The developed system captures a company's strategic information, provides facilities to quantify qualitative attributes and analyses them alongside the quantitative attributes in an evaluation framework. Through the system, similar cases can be retrieved to enable managers to make effective use of their knowledge and experience of previously delivered technologies and projects as an input to the prioritization of future projects. Other salient features of the system include its ability to adapt and absorb new knowledge and responses pertaining to significant events in the business environment, as well as to extract and elucidate information from the knowledge database for explaining and justifying its analysis. The applicability of the developed system is evaluated using a real case study in collaboration with a pharmaceutical manufacturing firm.