Abstracts :: David De Roure

< back to program

The myExperiment approach to Open Science

David De Roure*, Carole Goble**, Jiten Bhagat**, Don Cruickshank*
*University of Southampton, UK
**The University of Manchester, UK

myExperiment is a social web site for scientists which makes it easy to find, use and share scientific workflows and other scientific research objects, and to build communities [1]. In contrast to efforts which simply provide a repository or catalogue to share digital artefacts, myExperiment provides the social infrastructure to facilitate and encourage sharing. We believe this is an essential step in enabling Open Science to flourish: the http://www.myexperiment.org public site was made available in November 2007 and by May 2008 the site gained 850 registered users, with many others accessing
public content.

myExperiment aids reuse of digital assets because they can be discovered not just by what they do but based on how they are used by the community, with tags and reviews adding to the ‘collective intelligence’. However, several features distinguish myExperiment from other social web sites such as Facebook and MySpace. Firstly it is designed in very close cooperation with scientists and pays due attention to issues of ownership, attribution and licensing. Secondly it deals not in just one content type (photos, videos, slides, …) but supports the sharing of compound research objects which include distributed data, results, provenance information, tags, associated documentation, etc – reflecting the reality of experimental practice. It can exchange these objects using the emerging Object Reuse and Exchange standard from the Open Archives Initiative.

Significantly, myExperiment is open infrastructure for open science.  It is an open source codebase, implemented using the Ruby on Rails open source Web application framework using an agile and user-centric development method [2]. Individuals and laboratories are free to install their own myExperiment instances; they can then link them up using a federation model if and as they wish. While the public site provides a service for those who do not already have sharing and collaboration mechanisms in place, we also expose myExperiment functionality through simple RESTful APIs so that it can be accessed through existing interfaces, including wikis and web pages. This also enables the creation of other interfaces such as Google Gadgets, myExperiment add-ons for sites such as Facebook and functionality mashups over myExperiment. myExperiment could, for example, form the foundation of a personal or laboratory workbench.

The initial user community of myExperiment largely consists of scientific workflows users – especially bioinformaticians using the Taverna Workflow Workbench. Increasingly the project is also working with other disciplines, including chemistry and social science, and with other workflow systems. Workflow design is challenging and labour-intensive, and reuse is a particular challenge when scientists are outside a predefined Virtual Organisation or enterprise: rather they
are individuals or small groups, decoupled from each other and acting independently, who are seeking workflows for reuse or repurposing. myExperiment makes it easier for these workflow workers to gossip about and exchange workflows, and for inexperienced scientists to leverage the expertise of others. Furthermore, since the focus is on the workflow’s function and the services it uses, and not on the particular workflow engine, myExperiment is perhaps the first example of a multiworkflow environment for open science, where scientists can bring together workflows from
multiple systems.

MyExperiment is already forming the basis of other open science activities. For example, the myGrid project, along with the EBI, launched the BioCatalogue project in May 2008. This project builds on myExperiment to create a fully curated catalogue of Web services covering their functional, operational, usage, and provenance metadata. By analysing the designs, patterns and performance of services in workflows we can also improve the curation of services. In general, myExperiment supports open science by bringing the power of the community and network effects to tackle problems – such as reliability, curation and trust  – in ways that were just not possible before.

References:

[1] David De Roure, Carole Goble and Robert Stevens. Designing the myExperiment Virtual Research Environment for the Social Sharing of Workflows. e-Science 2007 – Third IEEE International Conference on e-Science and Grid Computing, 2007. Bangalore, India, 10-13 December 2007. Pages 603-610.
[2] De Roure, D. and Goble, C. (2007) Six Principles of Software Design to Empower Scientists. IEEE Software (to appear). See http://eprints.ecs.soton.ac.uk/15032/

< back to program

Advertisements

Responses

  1. […] Abstracts :: David De Roure […]

  2. […] Abstracts :: David De Roure […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: