Edinburgh Data-Intensive Research - UG4 student project

Improved data logging, sharing and analysis for the British Geological Survey's School Seismology project

Jano.van.Hemert — Mon, 23 Nov 2009 19:10:13 +0000

Student:

Jon Gilbert

The School Seismology project (http://www.bgs.ac.uk/schoolseismology/) enables schools to detect signals from large earthquakes happening anywhere in the world. It is used to teach a range of basic science concepts in over 400 schools around the UK by detecting world earthquakes in the classroom using a simple seismometer system and exchanging Earthquake data with schools around the world.

At the moment, the software (AmaSeis http://pods.binghamton.edu/~ajones/AmaSeis.html) that is used to monitor and analyse data from the seismometer systems is rapidly becoming out of date. It has several problems: it runs only on Windows, its server-client communication is hampered by certain firewalls, it requires a PC running 24/7, which costs about £200 in electricity per year, data has to uploaded to the sharing web site manually and data can only be analysed on computer that have AmaSeis installed.

In this project, you will develop a better way of handling data logging, sharing and analysis. Data logging must be done low-key so that it could run on cheap and low-powered devices (such as the NSLU2 http://en.wikipedia.org/wiki/NSLU2). Data sharing must be automatic by buffering and placing it into a central store at the British Geological Survey (BGS), taking into account whether an earthquake has happened and could have been recorded. The sharing must be via a protocol most firewalls will allow (HTTP). The analysis can be done with more modern software that already exists, such as SeisGram2k (http://alomax.free.fr/seisgram/SeisGram2K.html). The goal is to embed these into a simple portal which then runs these analysis on computers at the BGS, perhaps via Rapid (http://research.nesc.ac.uk/rapid/) if that is appropriate, in a similar way to RapidSeis (http://research.nesc.ac.uk/node/492).

You will be given access to a seismometer system and the current software. The primary aim is to replace the current bridge between the seismometer and the data sharing and analysis components. In combination with the BGS you will build a system that integrates with their current web site for sharing earthquake events and take their guidance in which analysis software to adopt.

This project may be eligible for a small honorarium under the BGS BUFI scheme http://www.bgs.ac.uk/research/bufi/home.html

Supervisors @ NeSC:

Jano.van.Hemert

Project status:

Finished

Degree level:

UG4

Other supervisors:

Paul Denton, British Geological Survey

Background:

Knowledge of Linux essential. Experience with web service development useful.

Subject areas:

e-Science

Computer Communication/Networking

Software Engineering

WWW Tools and Programming

Student project type:

UG4 student project

Predicting earthquake impact with laptop motion sensors

Jano.van.Hemert — Wed, 07 Jan 2009 21:24:06 +0000

Student:

Gary Mcgilvary

Grade:

first

Principle goal: to make clever use of the sudden motion sensor of Apple notebooks to map the impact of earthquakes in densely populated areas.

After an earthquake, it is vital to understand the damage sustained by its impact. Especially in densely populated areas, it is important for an effective emergency response to know which areas are likely to be worst affected. Predicting the damage caused by earthquakes is very difficult as it depends on many and poorly constrained factors from wave propagation characteristics to building vulnerability. Typically, the impact of earthquakes is linked to the level of shaking itself measured by sensors that pick up the ground acceleration [1]. To get an accurate picture of this, dense and then expensive networks are required which can only be applied in high hazard areas such as Tokyo or Taiwan for example.

In this project, the aim is to measure the impact of earthquakes by making use of the [2] sudden motion sensor found in recent Apple notebooks. These sensors are not as accurate as seismic sensors, but still sufficiently accurate to provide a map of impact providing many of them are distributed over the affected area. As cities are densely populated with many casual and business users, the chance of sufficient coverage is high. Also, as cities are often the scene for large amounts of damage by earthquakes, they provide the best application area for this approach.

To allow the approach to work, and to evaluate it, you will build a system based on existing prototypes [3,4]. The main idea is to have two components. First, a tool that people can install easily and which lets an Apple laptop continuously monitor "seismic" activity. The data produced must be buffered on the laptop for a limited period. Second, a system that when prompted with an earthquake queries the laptops it knows about (or even better, which are in the affected area) for their buffered data, which can then be copied to a central location for analysis. The output of this analysis can then be shown using so called felt-maps as shown in Figure 2 of [2].

By having your tool installed on multiple or even many laptops, you will then orchestrate your own "earthquake" by having everyone "shake" their laptops while recording the event. You will then create a ShakeMap, perhaps even showing the result on Google Maps?

This work will be in conjunction with the European-Mediterranean Seismological Centre [5], which primary purpose is to provide information on earthquakes in realtime. They will assist with the seismological part of your project including the analysis and visualisation.

Supervisors @ NeSC:

Jano.van.Hemert

Project status:

Finished

Degree level:

UG4

Other supervisors:

Rémy Bossu, European-Mediterranean Seismological Centre, France; Fabrice Cotton, Grenoble University, France

Subject areas:

e-Science

Distributed Systems

Mobile Computation

WWW Tools and Programming

Student project type:

UG4 student project

References:

[1] http://en.wikipedia.org/wiki/Seismic_wave [2] http://en.wikipedia.org/wiki/Sudden_Motion_Sensor [3] http://www.agu.org/pubs/crossref/2008/2008EO250001.shtml [4] http://www.suitable.com/tools/seismac.html [5] http://www.emsc-csem.org/

Mining and visualising family tree networks from online genealogy information

Jano.van.Hemert — Tue, 06 Jan 2009 16:29:28 +0000

Principle goal: to construct a data harvesting system with an associated semantic web-enabled store for genealogical data with a method for querying the data which you test using at least one query.

In the past decade there has been a massive increase of interest in genealogy due in part from the ready availability of much genealogical data via the web. One of the most important, and freely available, resources of genealogical information is provided by the Church of the Latter Day Saints [1]. The main aim of this project is to develop a system to automate the extraction and facilitate the analysis and visualisation of Mormon GEDCOM [2] data. In particular, the focus will be on Scottish genealogical data. The data will be stored in a format that can be readily used in programs devoted to spatial analysis such as graphical information systems (GIS). The benefits of having information in such a format will appreciated by a large number of fields, e.g. health, history, social science and economics to name just a few. This project will involve building a data harvesting system, developing an ontology for genealogical data and developing a structure in which to store the harvested data (e.g. in semantic web compatible format).

The analysis and visualisation phase of the project aims to build family tree networks from the extracted data in such a way it allows information about the size topology and shape of these trees to be extracted easily. By mapping this information to other data via GIS systems, it can then be put into new contexts. This allows geoscientists to answer question as to how the topology of family trees is affected by environmental issues, wars and epidemics. If you have sufficient time, you may try to answer one such question: “Using data from the Orkney islands, what are the main lines of communications between families on the main and more remote islands (i.e., the connectivity in the network)?” The results of this question could be then displayed on a simple GIS system such as DIY Map [3].

Supervisors @ NeSC:

Jano.van.Hemert

Project status:

Finished

Degree level:

UG4

Other supervisors:

Rachel Walcott (School of GeoSciences)

Background:

Applied Databases; Data Integration and Exchange; Querying and Storing XML; Knowledge Modelling and Management courses are desirable but not critical for this project.

Subject areas:

e-Science

Databases

WWW Tools and Programming

Student project type:

UG4 student project

References:

[1] http://www.familysearch.org [2] http://en.wikipedia.org/wiki/GEDCOM [3] http://www.backspace.com/mapapp/

Generating Graphical User Interfaces for Grid Computing Portals

Jano.van.Hemert — Mon, 04 Feb 2008 16:51:29 +0000

Student:

Carl Orebäck

Grade:

first

Carl was awarded both the Agilent Prize and the Class Medal for best performance in the BEng Electronics and Computer Science degree

To investigate the use of graphical user interface languages and their generators and use these in conjunction with the Rapid Development Tool for Job Submission Portlets (RAPID), which is being developed at the National e-Science Centre.

Grid computing has become very popular in the last decade with its promise to deliver unprecedented computing power to anyone on the planet, anywhere. However, to make use of any Grid, a number of software components and authentication systems are required. Typically, these components and systems vary from Grid to Grid, and without little exception, they are command-line based.

The command line is fine when working with computer-savvy scientists, such as physicists, engineers, astronomers and computer scientists. However, the majority of scientists will not have the time to invest in learning the command line, let alone the mystic incantations required to submit jobs and move data across Grids and other High Performance Computing facilities.

At the National e-Science Centre, a new tool called RAPID is under development that allows a rapid deployment of portals, i.e., web interfaces, that are specifically tailored to a group of scientists. Such a portal will allow them to perform tasks through filling in web forms and pressing the 'submit to Grid' button at the end.

The tool binds a user interface (xhtml forms) to an XML definition that specifies how a computer job will run on a Grid, called a Job Submission Description Language (JDSL). Currently, an extension is being developed that allows binding a database to both the xhtml forms and the JSDL. This binding consists of one XML file called RAPID-XML. It specifies the control flow of the user interface and in which way GUI components are connected to the JSDL file. No programming is required in the process.

Using the tool we can create and deploy portals quickly. However, this still needs a person skilled in the art of writing XML files. In this project you will investigate the use of languages and generators that can generate such an XML file by means of its own GUI. In other words, the aim is to have a user interface that generates user interfaces specifically for submitting compute jobs to Grids.

A suggested language and generator to look at is JAXFront/JAXForms.

Supervisors @ NeSC:

Jano.van.Hemert

Jos.Koetsier

Project status:

Finished

Degree level:

UG4

Background:

You should definitely not be afraid of XML.

Subject areas:

e-Science

Human-Computer Interaction

Other

Software Engineering

Projects:

Rapid

Student project type:

UG4 student project

References:

The RAPID project http://research.nesc.ac.uk/node/61

Association Rules Applied to Microarray Results from a Cystic Fibrosis Study

Jano.van.Hemert — Mon, 04 Feb 2008 16:27:02 +0000

Student:

Andrei Lyashko

Grade:

first

Note: Andrei's thesis was awarded a 91% and received the Best Undergraduate Project award sponsored by Microsoft

To perform data mining in the form of association rules on scientific data from a microarray study on Cystic Fibrosis, with the objective of evaluating and improving the mining algorithm.

The cystic fibrosis group in the Molecular Medicine Centre (MMC) at the Western General are developing gene gene therapies to treat cystic fibrosis, a life-threatening hereditary disease. To understand which genes are responsible for the disease, and which genes may have diagnostic or prognostic value, they go through a long process of laboratory experiments and analysis, involving microarray experiments and several types of statistical tests. A system is now in place that captures the data from a study that aims to find out which genes may be good targets for therapy into a standard (MySQL) database.

In this project, you will take the data from the target study and apply association rules, a particular type of data mining, to it. The aim is to find relationships between genes that help predict which processes should be changed in order to stop or reduce Cystic Fibrosis from affecting a patient. The advantage for the biologists is that they may discover more complex relationships, which cannot be found using the typical statistical procedures they currently depend on. This is an exciting project as your will work with real scientific data.

The scope of the project includes testing various derivations of the original association rules algorithms, such as extending it with quantifications of gene expression levels, and clustering within rules to understand how rules vary between patients and non-patients. Our first hypothesis is that we can find more interesting gene interaction patterns using association rules. Other hypotheses can be investigated along the lines of comparing different variants of the association rule mining algorithm.

You will work closely with three people, Jano van Hemert (National e-Science Centre) will provide support on data mining, Rob Kitchen (National e-Science Centre) will provide support on the microarray data management, and Varrie Ogilvie (Molecular Medicine Centre) will provide support on interpretation of microarray data.

Supervisors @ NeSC:

Jano.van.Hemert

Rob.Kitchen

Project status:

Finished

Degree level:

UG4

Other supervisors:

Varrie Ogilvie, Molecular Medicine Centre, University of Edinburgh

Subject areas:

e-Science

Algorithm Design

Bioinformatics

Machine Learning/Neural Networks/Connectionist Computing

Other

Student project type:

UG4 student project

References:

Non-technical explanation of association rules on Wikipedia http://en.wikipedia.org/wiki/Association_rule_learning Cystic Fibrosis on Wikipedia http://en.wikipedia.org/wiki/Cystic_fibrosis Microarray on Wikipedia http://en.wikipedia.org/wiki/DNA_microarray

GridDrive: Using sshfs to access the Grid

Jano.van.Hemert — Mon, 27 Aug 2007 09:50:38 +0000

Student:

Ciaran Hearne

Grade:

first

To patch sshfs to make it work with Grid certificates, then install it on a Linux machine and mount a directory on a machine that is part of the National Grid Service.

The Grid is a heterogeneous resource of machines that offer data storage and computing facilities. Although much research has gone into how to make interaction with computing facilities transparent to the user, this is not the case for the handling of data transfers.

In this project you will modify sshfs, a kernel extension that allows mounting directories over ssh, and FUSE, that allows users to mount discs in user-space. The aim is to patch sshfs in the same manner as ssh was patched to work with Grid certificates. These certificates are issued by parties involved in the National Grid Service, such as the National e-Science Centre, and allows users access to resources in many universities.

Once sshfs runs, anyone with a Grid certificate would be able to mount a directory "on their desktop" pointing to a machine that is part of the Grid. The next step is then to run a program that continuously looks for data files and then runs compute jobs on them.

This project requires a person with a hacker personality in the positive sense (http://en.wikipedia.org/wiki/Hacker), with much experience with Linux and its surrounding culture.

Supervisors @ NeSC:

Jano.van.Hemert

Jos.Koetsier

Project status:

Finished

Degree level:

UG4

Background:

You must like hacking Linux kernels and understand the idea behind ssh authentication, and be a proficient C programmer.

Subject areas:

Computer Architecture

Distributed Systems

Other

Software Engineering

Student project type:

UG4 student project

References:

SSHFS & FUSE http://fuse.sourceforge.net/sshfs.html National Grid Service http://www.grid-support.ac.uk/ x509 proxy certificates http://www.ietf.org/rfc/rfc3820.txt x509 patch for SSH http://roumenpetrov.info/openssh/