Waikato University crest

Department of
Computer Science
Tari Rorohiko

Computing and Mathematical Sciences

Recent Seminars

Events Index

The information technology-autism connection

Antony Thomas
Clinical Psychologist & Behaviour Analyst, NZ Ministry of Education, Special Education, Hamilton
Tuesday 10 December 2013
There are two prevalent, but contradictory, stereotypes on autism; the first one is of a non-communicative person who engages in self stimulatory/weird behaviours, and the other one is of a person who exhibits many of the characteristics of a creative genius and of a brilliant person. The movie, “Rain Man” might have largely contributed to some of these stereotypes.

The landscape of autism has changed drastically in the past three decades. We now know that a person with High Functioning Autism can engage in any kind of job /profession, provided he has an interest in that activity and can lead a normal or near normal social life. Some of the famous and brilliant people who showed signs of autism include Albert Einstein, Isaac Newton, George Orwell, H.G wells. Beethoven, Mozart, Hans Christian Anderson, and Thomas Jefferson.

Some people on the autism spectrum possess superior skills. The ability to focus on a given task for longer periods of time can be an advantage for certain professions. A 2012 study showed that people with autism possess greater than normal capacity for processing information. The research may help to throw light on the higher than average prevalence of people with autistic features in the IT industry. It is not a secret that some of the IT companies are showing extra preference for recruiting people on the autism spectrum for certain type of IT jobs.

It is also a common observation that children with autism are attracted to computers from a very young age. Some of them are even addicted to computers and technology. People with autism prefer to interact with computers more often than to interact with humans because computers are pretty predictable whereas human beings are usually not.

Dr. Thomas, will discuss further on the personality characteristics of autism and the unique dynamics involved in the Information technology - autism connection and attraction. Some the recent research findings on this topic will also be highlighted.

 

What can we learn from software failure data?

Veena Mendiratta
Bell Labs
Friday 29 November 2013
Failure detection and fault correction are vital to ensure high quality software. During the development and deployment phases detected failures are commonly classified by severity and tracked to meet quality and reliability requirements. Besides tracking failures, this data can be analyzed and used to qualify the software and to control the development and maintenance process. Our work is focused on failure data collected during the development phase and explores what we can learn by analyzing this data. Change management systems log the failures detected and the code fixes to correct the underlying software defects. By applying software reliability models and statistical techniques to this defect data, we can answer questions such as the following:
  • Is the maintenance process increasing the software reliability?
  • Is the maintenance process under control?
  • How many failures are expected to occur in the field?
  • What is the expected time remaining to meet the reliability requirement?

This presentation addresses these questions by using a methodology based on trend analysis, control charts and software reliability growth models. The methodology is applied to a large software system during various stages of testing including customer acceptance testing. What is new about this methodology is the combined use of control charts, trend analysis and software reliability models.

 

Avarit & Amharic: Combining Hebrew, Arabic, Latin and Ethopicinto a new writing system

Liron Turkenich
AUT University, Auckland
Friday 1 November 2013
Aravrit presents a set of hybrid letters merging Hebrew and Arabic. This new writing system is composed of an Arabic letter on the upper half and a Hebrew letter on the bottom half. The characteristic features of each letter were retained, however in both languages the fusion required some compromises to be made, yet maintaining readability and with limited detriment to the original script. In Aravrit, one can read the language he/she chooses, without ignoring the other one, which is always present. Liron's new project is Amharic which combines Latin, Hebrew and Ethiopic scripts.

 

Branding, society and popular culture

Gjoko Muratovski
AUT University, Auckland
Thursday 24 October 2013
In our post-industrial society, consumer goods have transcended the realm of material things and have become a central idea of everyday living. In ever-increasing numbers, people purchase products, use services, or interact with organisations that help them to express themselves, or to feel more like who they want to be, in their own eyes and in the eyes of their peers. Products and businesses have taken abstract and conceptual forms around which a meaningful existence is formed through the use of branding. In line with these developments the focus within the field of design has been shifting from functional and technical components to narrative components. Design has led itself on a search through both design history and material culture for elements of our physical vocabulary that can be called upon anew. Branding has undergone similar processes of self-discovery. Commercial brands have become containers for stories, and correspondingly, designers, brand managers, advertisers, and ‘spin doctors' have become storytellers, and even mythmakers. This seminar will examine branding not merely as a marketing tool, but as a socio-cultural phenomenon that is central to our culture.

 

PAN agency: A case study on PAN's approach to a unique social health challenge, STI's

Toby Hutchings
Digital Director, PAN
Wednesday 23 October 2013
PAN® provides integrated marketing strategies that bring together marketing disciplines from design to web development, brand to campaign management, resulting in meaningful outcomes across multiple consumer touch-points. In the past this required multiple agencies, consultants and suppliers plus many additional layers of resource and communication channels. As a hybrid agency, PAN effectively cuts through the clutter to deliver outstanding results-oriented marketing solutions.

PAN® will be presenting their hybrid agency direction, discussing how and why the blend of creative, campaign, online and web works for them and those around them. Toby and Graeme, along with multi disciplined creative designer Hemi Ormsby, will be on hand to present and discuss a case study on PAN's approach to a unique social health challenge relevant to students and faculty alike: STI's.

 

Web Dev, Mobile, HPC and friends – Environmental Informatics at Landcare Research

Markus Müller
Landcare Research, Hamilton
Tuesday 22 October 2013
Environmental informatics is the science of information applied to environmental science. As such, it provides the information processing and communication infrastructure to the interdisciplinary field of environmental sciences aiming at data, information and knowledge integration. An important aspect of environmental data and therefore Environmental Informatics is its geospatial nature.

One of seven Crown Research Institutes (CRIs), Landcare Research’s core purpose is to drive innovation in the management of terrestrial biodiversity and land resources. Environmental Informatics is an integral part of many of the research projects carried out.

The seminar will showcase some exemplar projects thereby describing the overall scope of Environmental Informatics at Landcare Research. The examples are grouped around the themes web development, mobile applications and High Performance Computing (HPC).

 

Mining the radio sky for pulsars and transients

Willem van Straten
Swinburne University of Technology, Melbourne
Friday 18 October 2013
The radio sky is buzzing and clicking with signals from some of the most exotic and mysterious phenomena in the Universe. However, as more and more of the radio spectrum is polluted by our own devices, it is becoming increasingly difficult to hear the distant astronomical signals buried in the interference. It is no longer possible for a small team of researchers to sift through the haystack of noise in search of the physically significant needle; therefore, scientists have turned to “citizen science” and “artificial intelligence” as means of classifying large data sets in a reasonable amount of time. I will review some of the recent progress in this area, discuss the motivation for further research and development, and consider the new challenges posed by the next generation of radio observatories, such as the Square Kilometre Array.

 

Studies at NUS: Gait and PAT

Associate Professor Hugh Anderson
SoC, National University of Singapore
Tuesday 24 September 2013
A system for monitoring the gait of Alzheimer's patients has a number of distinct components, with different types of communication between each component type. This motivates the use of tools like PAT, the Process Analysis Toolkit, which attempts to bring together a range of tools and notations for modelling complex parallel systems, and then checking those models for desirable properties. PA attempts to make model checking easy to use and apply, providing a range of graphical and program-style interfaces, to assist you in the development of complex models. The different interfaces help convince you that your model really does model something and that the property you are checking is relevant. One of the features of PAT is an emphasis on using the underlying model checkers in somewhat unusual ways. For example, using the counterexample traces to solve shortest path problems. At the end of the talk, Hugh will outline the possibilities of research at the School of Computing at NUS. This should be of particular interest to any students who are contemplating doing their PhD studies outside of NZ.

 

Māori typeface design and Māori typography

Johnson Witehira
Kaupapa Māori Graphic Design
Wednesday 11 September 2013
Johnson will be giving a talk on Māori typeface design and Māori typography - this is centered around the process and research he went through in developing his typeface design. Johnson Witehira is an artist, designer and researcher of Nga Puhi and Tamahaki (Whanganui) descent. In his recently completed doctoral thesis, he looked at how customary carving might provide a model for contemporary Māori design practice.The research responds to the dearth of Māori informed guidelines for designers, both Māori and non-Māori, when working with Māori content, form and imagery. In view of the increased use of Māori iconography in design industries both locally and globally, there is a need to develop guidelines that help maintain the integrity and intent of the Māori form and content, while enabling designers to express culturally significant messages. Through this research he also sought to reveal a uniquely Māori design language, and establish the Māori elements and principles of design.

 

Designing with the mind in mind: the psychological basis for UI design rules

Jeff Johnson
Principal Consultant at UI Wizards, Inc.
Thursday 5 September 2013
UI design rules are not simple recipes to be applied mindlessly. Applying them effectively requires determining their applicability and precedence in specific situations. It also requires balancing the trade-offs that inevitably arise in situations when design rules appear to contradict each other. By understanding the underlying psychology for the design rules, designers and evaluators enhance their ability to interpret and apply them. Explaining that psychology is the focus of this talk. It is intended mainly for software designers and developers who did not take Cognitive Psychology as part of their university education, or who took it long ago and would like an update based on recent research. This talk is based on Jeff Johnson's recent book of the same title.

 

From Waikato to Microsoft

Mark S Staveley
Microsoft
Tuesday 6 August 2013
Have you ever wondered what is in store for you after your Computer Science studies? Do you wrestle with trying to decide between academia, research, and industry? Do you think about perhaps starting up your own company? Or are you more interested in joining a big company such as Microsoft?

Please come and join Mark Staveley (Waikato CS MSc Grad ’00) as he shares some of the interesting experiences and lessons learned over the past 13 years since he left Waikato.

 

Interactive learning of dependency networks for scientific discovery

Diane Oyen
University of New Mexico
Tuesday 21 May 2013
Machine learning algorithms for identifying dependency networks are being applied to data in biology to learn protein correlations and neuroscience to learn brain pathways associated with development, adaptation and disease. To ask comparative questions, multiple networks must be considered simultaneously; exploding the hypothesis space of the learning problem. Exploring these complex solution spaces requires input from the domain scientist to refine the objective function. In this talk, I present transfer learning algorithms for both Bayesian networks and graphical lasso. By incorporating human input in the transfer bias objective, the topology of the solution space is shaped to answer knowledge-based queries about the confidence of dependency relationships. I also describe an interactive human-in-the-loop approach that allows a human to react to machine-learned solutions and give feedback to adjust the objective function. The result is a solution to an objective function that is jointly defined by the machine and a human.

 

Unlocking the secrets of 3 billion pages: Introducing the HathiTrust Research Center

Stephen Downie
University of Illinois at Urbana-Champaign
Wednesday 15 May 2013

This briefing will introduce the HathiTrust Research Center (HTRC). The HTRC is the research arm of the HathiTrust, an online repository dedicated to the provision of access to a comprehensive body of published works for scholarship and education. The HathiTrust is a partnership of over 60 major research institutions and libraries working to ensure that the cultural record is preserved and accessible long into the future. HathiTrust membership is open to institutions worldwide.

Over 10 million volumes have been ingested into the HathiTrust digital archive from sources including Google Books, member university libraries, the Internet Archive, and numerous private collections. The HTRC is dedicated to facilitating scholarship using this enormous corpus through enabling access to the corpus, developing research tools, fostering research projects and communities, and providing additional resources such as enhanced metadata and indices that will assist scholars to more easily exploit the HathiTrust corpus. This briefing will outline the mission and goals of the HTRC. It will also introduce current and planned projects, including its work on enabling the non-consumptive analyses of copyrighted materials. It will conclude with a discussion of the ways in which scholars can work with and through the HTRC.

 

Policy search based relational reinforcement learning using the cross-entropy method

Sam Sarjant
Department of Computer Science, The University of Waikato
Thursday 2 May 2013
In the field of relational reinforcement learning — a representational generalisation of reinforcement learning — the dynamic nature of environments results in a potentially infinite number of possible states, requiring learning agents to use some form of abstraction to learn effectively. Instead of forming an abstraction over the state-action space, an alternative technique is to create behaviour directly through policy-search. The algorithm named CERRLA presented in this talk uses the cross-entropy method to learn behaviour directly in the form of decision-lists of relation rules for solving problems in a range of different environments, without the need for pre-defined abstractions or guidance. The behaviour produced by the algorithm is easy to comprehend and is biased towards compactness. The results obtained show that CERRLA is competitive in both the standard testing environment and in Ms. Pac-Man and Carcassonne, two large and complex game environments.

 

On compositional supervisor synthesis for discrete event systems

Sahar Mohajerani
Department of Signals and Systems at Chalmers University of Technology, Sweden
Tuesday 30 April 2013
Supervisory control theory provides a general framework to automatically calculate control functions for discrete event systems such as manufacturing systems, embedded functions etc. Calculating such a supervisor for large systems stumble on an inherent complexity problem known as the state-space explosion problem. One method to alleviate the state-space explosion problem is compositional approach, which exploits the modular structure of a system to reduce the size of the system.

 

The Whiley programming language: design & implementation

David Pearce
Victoria University of Wellington
Friday 26 April 2013
Whiley is a new programming language being developed at Victoria University of Wellington, New Zealand (see http://whiley.org). Whiley is designed specifically to simplify program verification. This goal has constrained the language in many ways, some of which will be expected whilst others are less apparent. The language includes first-class pre- and post-conditions, and the aim is to automatically check them at compile time. Such constraints must be pure and may range over first-class data-types (sets, lists, etc) with value semantics, and also functions that are explicitly declared pure. To simplify verification, arithmetic operates over unbounded integers and rationals. For flexibility, a flow-sensitive type system with structural subtyping is employed.

Despite being a primarily functional language, Whiley retains a distinctly imperative syntax with the look and feel of a dynamic language (i.e. Python). Whiley currently compiles to the JVM and is fully inter-operable with Java. Many challenges exist in compiling Whiley programs to the JVM for efficient execution. Furthermore, other compilation targets (e.g. JavaScript) are planned for the future. In this talk, I will introduce and demonstrate Whiley, as well as discuss some of the main challenges involved.

 

What is an ORCID and why should I get one?

Helen Lynch
Library, The University of Waikato
Tuesday 23 April 2013

ORCIDs provide a way of distinguishing your research activities from those of others with similar names and to attach your unique identity to research outputs.

Open Researcher and Contributor ID (ORCID) is an open, non-profit organisation that maintains an international registry of unique researcher identifiers and a method of linking research activities to those identifiers. You can link to your other identifiers (such as Scopus Author ID or ResearcherID or LinkedIn). You can include your ORCID ID on your webpage, when you submit publications, apply for grants, and in any research workflow to ensure you get credit for your work. This presentation will show you how to set up your own ORCID.

http://orcid.org/about/what-is-orcid

 

Bigraphical reactive systems for domain-specic modelling languages

Gian Perrone
Tuesday 19 March 2013
Modelling is a ubiquitous activity in human endeavours, and the construction of informatic models of many kinds is the key to understanding and managing the complexity of an increasingly computational world. We advocate the use of domain-specic modelling languages, instantiated within a "tower" of models, in order to improve the utility of the models we build, and to ease the process of model construction by moving the languages we use to express such models closer to their respective domains. This talk is concerned with the study of bigraphical reactive systems as a host for domain-specic modelling languages. We present a number of novel technical developments, including a new complete meta-calculus presentation of bigraphical reactive systems.

 

Mining moving objects: challenges and directions

Dino Ienco
Irstea Institute, Montpellier, France
Tuesday 19 February 2013
Nowadays, the use of many electronic devices in real world applications has led to an increasingly large amount of data containing moving object information. One of the objectives of spatiotemporal data mining is to analyze such datasets to understand the object movement behavior. In this context, many recent studies have been defined to mine moving object clusters including flocks, moving clusters, convoy queries, closed swarms, group patterns, etc. Although these patterns are interesting and useful, mining and analyzing them is still challenging. There are several key questions, to answer, such as:
  • How can we extract and compare them efficiently?
  • Are they good enough to represent the complex object movement behavior?
  • Among the thousands of patterns, which ones are the most important to summarize a moving object dataset?
In this talk, we introduce, analyze and address these three challenges and propose future research directions.

 

OpenPAT: Analysing programs the easy way — preview of invited presentation at Victoria

Simon Spacey
Department of Computer Science, The University of Waikato
Friday 15 February 2013
OpenPAT.org is the home of the Open Program Analysis Toolkit project. OpenPAT differs from other program analysis toolkits in that it instruments code statically and gathers dynamic timing, control and data flow information as the instrumented program runs. In this presentation we will introduce the new OpenPAT toolkit (version 3), review the OpenPAT approach and examine its benefits in comparison with alternative toolkits and then we will create a new tool for OpenPAT with just a few lines of code that can be used to analyse the internal workings of programs written in any compilable language.

 

Predicting the rankings of financial analysts: an application of label ranking methods

Carlos Soares
University of Porto, Portugal
Tuesday 12 February 2013
The goal of this work it to develop trading strategies based on the predicted performance of financial analysts. In other words, we want to predict which analysts will make accurate recommendations concerning whether to buy or sell a stock. We assume that predicting the performance of these analysts is less complex than predicting the behavior of the markets. Given that in a given period, 1) different analysts may make recommendations but 2) not all analysts are expected to make them, we need to predict their relative performance. Therefore, we address the problem as a label ranking task. In this talk, I will give an overview of our work on label ranking in the context of predicting rankings of financial analysts and other applications.

 

Events Index