5,670 results for Conference item

  • Developing an online learning community: A model for enhancing lecturer and student learning experiences

    Khoo, Elaine G.L.; Forret, Michael; Cowie, Bronwen (2009)

    Conference item
    University of Waikato

    This paper reports on a study aimed to better understand teaching and learning in an online learning environment through the development of a learning community to facilitate successful learning experiences. To achieve this aim, a qualitative interpretive methodology was adopted to case study an online lecturer and his 14 students’ experiences in a semester long fully online asynchronous graduate course in a New Zealand tertiary institution. Based on the findings, a model for understanding and developing an online learning community for adult tertiary learners is proposed. In accord with sociocultural views of learning and practices, the model depicts successful online learning as a mediated, situated, distributed, goal-directed and participatory activity within a socially and culturally determined learning community. The model informs our understanding of appropriate conditions for the development of online learning communities and has implications for the design and facilitation of learning in such contexts.

    View record details
  • It only took 2 clicks and he’d lost me: Dimensions of inclusion and exclusion in ICT supported tertiary engineering education

    Khoo, Elaine G.L.; Johnson, E. Marcia; Torrens, Rob; Fulton, Jim (2011)

    Conference item
    University of Waikato

    Current conceptualisations of the digital divide have broadened beyond the notion of ‘haves’ and ‘have nots’ to include a more multifaceted perspective in which individuals and the contexts in which they learn are explicitly considered. This paper reports on a qualitative case study of a compulsory Engineering foundations course at a tertiary institution in New Zealand. The course provides a broad introduction to engineering concepts, with particular emphasis on problem solving, the design process, and use of 3-dimensional computer-aided design (CAD) software. Findings illustrate and illuminate the multidimensional nature of information and communication technology (ICT) inclusion/ exclusion and are described within three themes – technological, conceptual, and aspirational/ professional. Implications are presented for course designers and lecturers interested in providing more inclusive learning environments.

    View record details
  • “It gave me a much more personal connection”: Student generated podcasting and assessment in teacher education

    Forbes, Dianne Leslie; Khoo, Elaine G.L.; Johnson, E. Marcia (2012)

    Conference item
    University of Waikato

    This paper reports on a qualitative case study of an online initial teacher education class in New Zealand, exploring the potential of student-generated podcasts as a form of interactive formative assessment. Findings from interviews with teaching staff indicate that podcasting was useful for supporting multimodal learning valuing student voice and reflections. Podcasting enhanced the affective and relational connections in the online class, and empowered students to develop technical skills and confidence relevant in their teaching careers. As such, this study positions educators as future makers and as leaders in a climate of change. We suggest implications for student-generated podcasts in similar contexts.

    View record details
  • Learning threshold concepts in an undergraduate engineering flipped classroom

    Peter, Mira; Khoo, Elaine G.L.; Scott, Jonathan B.; Round, W. Howell (2016)

    Conference item
    University of Waikato

    Given that the current goals for tertiary education is to better prepare students to apply their disciplinary knowledge in the real world and novel situations, it is imperative that students master the necessary disciplinary threshold concepts and competencies. Building on the findings of our pilot study of a partly-flipped undergraduate electronic engineering course, a version of a fully flipped is implemented in an intensive six-week version of the course involving in-class collaborative problem solving and continuous assessment. Data collected from the 32 students enrolled in the course include student surveys, video analytics, weekly student assessments, class observations and a focus group interview. Although data collection is still underway, the emerging findings indicate that students are watching the recommended weekly videos prior to coming to class and are solving online tutorials problems much more diligently, resulting in higher levels of in-class student collaboration compared to the pilot study. The results are discussed in regard to the effects of the fully flipped class model and the continuous assessment on students’ learning of threshold concepts and competencies.

    View record details
  • Evaluating Flipped Classrooms with respect to Threshold Concepts Learning in Undergraduate Engineering

    Khoo, Elaine G.L.; Scott, Jonathan B.; Peter, Mira; Round, W. Howell (2015)

    Conference item
    University of Waikato

    This paper reports on the initial findings from a two year (2015-2016) investigation of the impact of the flipped classroom on student learning of threshold concepts (TCs) in a large introductory undergraduate engineering course at a New Zealand university. As part of the flipped class intervention trialed over a threeweek period, a series of short themed video lectures were developed as a replacement for the traditional weekly lectures. The weekly practical lab session were redesigned to incorporate small-group problem solving tasks and assessment. Data from student surveys, interviews, class observations, and video analytics were collected and analyzed. Findings revealed that students were familiar with online videos as a learning resource; they had positive past experiences with using them and were willing to participate in a flipped classroom. However, most students did not watch all assigned weekly videos, including ones crucial to their TC learning. There is indication they thought learning strategies involving interactions with real persons to be more useful to their learning. This suggests that current strategies for motivating students to access and engage with the prepared videos need to be revised to maximize students’ learning opportunities.

    View record details
  • Supporting primary student independence in virtual learning: Investigating the role of school-based support staff

    Whalley, Rick; Khoo, Elaine G.L. (2016)

    Conference item
    University of Waikato

    This paper reports on the emerging findings of a small qualitative study investigating the role of school based support staff (hereafter referred to as SBSS) in supporting students to become independent virtual learners in the Virtual Learning Network Primary School (VLNP). The VLNP is a collaboration of schools throughout New Zealand providing virtual learning opportunities for their students in subjects that are not available in their own schools. The SBSS are staff members in the student’s home school who support and mentor the student during their time in the VLNP. Students that learn through the VLNP have varying levels of academic, technical and independent skills. In some schools SBSS assist students, however the expectations and degree of support varies from school to school. Two schools within the VLNP were used in this study. An interpretive qualitative methodology was adopted using individual semi-structured online interviews with the teachers, eteachers, principals and students at each of the case study sites. Grounded Theory was used to analyse the data. Eight key themes emerged to highlight the multiple roles that the SBSS importantly play in the VLNP. These include developing critical thinking, providing a wrap around approach, removing barriers to learning, providing opportunities, tuakana/teina: learning from each other, allowing students to take responsibility for their own learning, monitoring teaching and learning, and having administrative/managerial processes in place. The key findings in this study are of distributed support by all stakeholders and the importance of the role of the SBSS in coordinating this support.

    View record details
  • Advanced Recommendation Models for Mobile Tourist Information

    Hinze, Annika; Junmanee, Saijai (2006)

    Conference item
    University of Waikato

    Personalized recommendations in a mobile tourist information system suffer from a number of limitations. Most pronounced is the amount of initial user information needed to build a user model. In this paper, we adopt and extend the basic concepts of recommendation paradigms by exploiting a user’s personal information (e.g., preferences, travel histories) to replace the missing information. The designed algorithms are embedded as recommendation services in our TIP prototype. We report on the results of our analysis regarding effectiveness and performance of the recommendation algorithms. We show how a number of limiting factors were successfully eliminated by our new recommender strategies.

    View record details
  • Evaluating the replicability of significance tests for comparing learning algorithms

    Bouckaert, Remco R.; Frank, Eibe (2004)

    Conference item
    University of Waikato

    Empirical research in learning algorithms for classification tasks generally requires the use of significance tests. The quality of a test is typically judged on Type I error (how often the test indicates a difference when it should not) and Type II error (how often it indicates no difference when it should). In this paper we argue that the replicability of a test is also of importance. We say that a test has low replicability if its outcome strongly depends on the particular random partitioning of the data that is used to perform it. We present empirical measures of replicability and use them to compare the performance of several popular tests in a realistic setting involving standard learning algorithms and benchmark datasets. Based on our results we give recommendations on which test to use.

    View record details
  • Thesaurus based automatic keyphrase indexing

    Medelyan, Olena; Witten, Ian H. (2006)

    Conference item
    University of Waikato

    We propose a new method that enhances automatic keyphrase extraction by using semantic information on terms and phrases gleaned from a domain-specific thesaurus. We evaluate the results against keyphrase sets assigned by a state-of-the-art keyphrase extraction system and those assigned by six professional indexers.

    View record details
  • Vehicle lead-acid battery state-of-charge meter

    Scott, Jonathan B.; Pennington, Kyle; Schwarz, Sergei; Rowe, Philip (2011)

    Conference item
    University of Waikato

    We describe a state-of-charge, or “residual-capacity” meter for lead-acid batteries that intelligently synthesizes coulometric and terminal-voltage methods in a new algorithm to provide reliable, continuous readout of remaining capacity. Novel electronic circuit design eliminates the need to install a shunt in the vehicle. The meter learns the characteristics of a battery to which it is attached, removing the need for setup, customisation, programming or calibration at time of installation or battery replacement. The meter can thus be installed by unqualified personnel. Initial measurements suggest the design to be robust and accurate.

    View record details
  • Beyond trees: Adopting MITI to learn rules and ensemble classifiers for multi-instance data

    Bjerring, Luke; Frank, Eibe (2011)

    Conference item
    University of Waikato

    MITI is a simple and elegant decision tree learner designed for multi-instance classification problems, where examples for learning consist of bags of instances. MITI grows a tree in best-first manner by maintaining a priority queue containing the unexpanded nodes in the fringe of the tree. When the head node contains instances from positive examples only, it is made into a leaf, and any bag of data that is associated with this leaf is removed. In this paper we first revisit the basic algorithm and consider the effect of parameter settings on classification accuracy, using several benchmark datasets. We show that the chosen splitting criterion in particular can have a significant effect on accuracy. We identify a potential weakness of the algorithm—subtrees can contain structure that has been created using data that is subsequently removed—and show that a simple modification turns the algorithm into a rule learner that avoids this problem. This rule learner produces more compact classifiers with comparable accuracy on the benchmark datasets we consider. Finally, we present randomized algorithm variants that enable us to generate ensemble classifiers. We show that these can yield substantially improved classification accuracy.

    View record details
  • Speeding up logistic model tree induction

    Sumner, Marc; Frank, Eibe; Hall, Mark A. (2005)

    Conference item
    University of Waikato

    Logistic Model Trees have been shown to be very accurate and compact classifiers [8]. Their greatest disadvantage is the computational complexity of inducing the logistic regression models in the tree. We address this issue by using the AIC criterion [1] instead of cross-validation to prevent overfitting these models. In addition, a weight trimming heuristic is used which produces a significant speedup. We compare the training time and accuracy of the new induction process with the original one on various datasets and show that the training time often decreases while the classification accuracy diminishes only slightly.

    View record details
  • Modelling safety properties of interactive medical systems

    Reeves, Steve; Bowen, Judy (2013)

    Conference item
    University of Waikato

    Formally modelling the software functionality and interactivity of safety-critical devices allows us to prove properties about their behaviours and be certain that they will respond to user interaction correctly. In domains such as medical environments, where many different devices may be used, it is equally important to ensure that all devices used adhere to a set of safety, and other, principles designed for that environment. In this paper we look at modelling important properties of interactive medical devices including safety considerations mandated by their users. We use ProZ for model checking to ensure that properties stated in temporal logic hold, and also to check invariants. In this way we gain confidence that important properties do hold of the device, and that models of particular devices adhere to the properties described.

    View record details
  • Evolving triggers for dynamic environments

    Trajcevski, Goce; Scheuermann, Peter; Ghica, Oliviu; Hinze, Annika; Voisard, Agnes (2006)

    Conference item
    University of Waikato

    In this work we address the problem of managing the reactive behavior in distributed environments in which data continuously changes over time, where the users may need to explicitly express how the triggers should be (self) modified. To enable this we propose the (ECA)2 – Evolving and Context-Aware Event-Condition-Action paradigm for specifying triggers that capture the desired reactive behavior in databases which manage distributed and continuously changing data. Since both the monitored event and the condition part of the trigger may be continuous in nature, we introduce the concept of metatriggers to coordinate the detection of events and the evaluation of conditions.

    View record details
  • Kava: Untangling fact from fiction

    Aporosa, Apo (2017)

    Conference item
    University of Waikato

    Kava, in both its plant and drink form, is Pasifika’s ‘cultural keystone species’ and a potent icon of identity with some of its medicinal efficacy legitimised within Western pharmacology and research. However, for every positive concerning kava there appears to be a counterpoint: kava is being ‘abused’; kava causes liver damage; kava encourages men to stay away from home for lengthy periods negatively impacting the family; kava turns it’s drinkers into Zombies incapable of functioning the next day, etc. This presentation addresses these claims while also seeking reasons as to what motivates kava criticism.

    View record details
  • Issues in Location-Based Indexing for Co-operating Mobile Information Systems

    Osborn, Wendy; Hinze, Annika (2007)

    Conference item
    University of Waikato

    Mobile information systems need to collaborate with each other to provide seamless information access to the user. Information about the user and their context provides the points of contact between the systems. Location is the most basic user context. TIP is a mobile tourist information system that also provides location-based access to documents in the digital library Greenstone. This paper identifies the challenges for providing efficient access to location-based information using the various access modes a tourist requires on their travels. We discuss our extended 2DR-tree approach to meet these challenges.

    View record details
  • Combining Naive Bayes and Decision Tables

    Hall, Mark A.; Frank, Eibe (2008)

    Conference item
    University of Waikato

    We investigate a simple semi-naive Bayesian ranking method that combine naive Bayes with induction of decision tables. Naive Bayes and decision tables can both be trained efficientyly, and the same holds true for the combined semi-naive model. We show that the resulting ranker, compared to either component technique, frequently significantly increases AUC. For some datasets it significantly improves on both techniques. This is also the case when attribute selection is performed in naive Bayes and its semi-naive variant.

    View record details
  • Fast conditional density estimation for quantitative structure-activity relationships

    Buchwald, Fabian; Girschick, Tobias; Kramer, Stefan; Frank, Eibe (2010)

    Conference item
    University of Waikato

    Many methods for quantitative structure-activity relationships (QSARs) deliver point estimates only, without quantifying the uncertainty inherent in the prediction. One way to quantify the uncertainy of a QSAR prediction is to predict the conditional density of the activity given the structure instead of a point estimate. If a conditional density estimate is available, it is easy to derive prediction intervals of activities. In this paper, we experimentally evaluate and compare three methods for conditional density estimation for their suitability in QSAR modeling. In contrast to traditional methods for conditional density estimation, they are based on generic machine learning schemes, more specifically, class probability estimators. Our experiments show that a kernel estimator based on class probability estimates from a random forest classifier is highly competitive with Gaussian process regression, while taking only a fraction of the time for training. Therefore, generic machine-learning based methods for conditional density estimation may be a good and fast option for quantifying uncertainty in QSAR modeling.

    View record details
  • Sentiment knowledge discovery in Twitter streaming data

    Bifet, Albert; Frank, Eibe (2010)

    Conference item
    University of Waikato

    Micro-blogs are a challenging new source of information for data mining techniques. Twitter is a micro-blogging service built to discover what is happening at any moment in time, anywhere in the world. Twitter messages are short, and generated constantly, and well suited for knowledge discovery using data stream mining. We briefly discuss the challenges that Twitter data streams pose, focusing on classification problems, and then consider these streams for opinion mining and sentiment analysis. To deal with streaming unbalanced classes, we propose a sliding window Kappa statistic for evaluation in time-changing data streams. Using this statistic we perform a study on Twitter data using learning algorithms for data streams.

    View record details
  • Speeding up and boosting diverse density learning

    Foulds, James Richard; Frank, Eibe (2010)

    Conference item
    University of Waikato

    In multi-instance learning, each example is described by a bag of instances instead of a single feature vector. In this paper, we revisit the idea of performing multi-instance classification based on a point-and-scaling concept by searching for the point in instance space with the highest diverse density. This is a computationally expensive process, and we describe several heuristics designed to improve runtime. Our results show that simple variants of existing algorithms can be used to find diverse density maxima more efficiently. We also show how significant increases in accuracy can be obtained by applying a boosting algorithm with a modified version of the diverse density algorithm as the weak learner.

    View record details