Stage 2 Grant

Proposal Summary


The British Museum (BM) requests an Officer’s Grant of $100,000, for use over 4 months, to support initial planning, analysis and feasibility work for the ResearchSpace initiative.   Such work would be undertaken with the understanding that should the planning phase prove successful, the BM would be invited to prepare a proposal for consideration by the Mellon board to support development and implementation of ResearchSpace. 

The ResearchSpace concept, fully described in the appended documents, encompasses a shared information technology infrastructure together with software tools that would facilitate and encourage the sharing of information and data across projects and organisations.  As envisaged, when fully developed, ResearchSpace would deliver the following elements:

  1. A Collaboration Environment – Infrastructure and software tools that facilitate two or more people or organisations working together to achieve common goals and objectives.
  2. Harmonised data – Museum collection, conservation, and scientific data stored in a semantic RDF form so that it can be shared and reused among projects without expensive integration software.
  3. Research tools – The provision of components that can be used independently or in combination with other components for data analysis and manipulation to support a particular collaborative workflow.

The present proposal is for funds to support the work needed to define and specify the resources required to initiate and complete a full implementation phase. The goal of this preliminary work is to provide, by the end of August 2010, a practical and costed design for a shared technology infrastructure, ready for procurement, development and implementation.  Such an infrastructure would have as its core goal to fulfil the needs of all of the Mellon Foundation’s Museum and Art Conservation’s (MAC) program prototypes to present machine readable collection, conservation, and scientific information via the internet.  The broader goal would be to build ResearchSpace in such a way as to support the storage and exposure of any relevant cultural heritage data and multimedia using semantic technologies.

Since RDF data services represent the least known element in this enterprise, the emphasis of the planning phase (and associated budget) is on mitigating the risk of this element, thereby reducing the possibility of technical delays during implementation. Successful testing would also provide the project team and sponsors with greater confidence in the relatively new technology approach. The results of this work should be reusable and reduce the timeline for a development effort.

Main Proposal

Background


The ResearchSpace concept is based on a proposal for a shared technology infrastructure provided by staff of the Research and Information Technology (RIT) Program at the Andrew W. Mellon Foundation. This proposal was circulated by staff of the Museums and Art Conservation (MAC) Program on August 25, 2009 for discussion at a meeting held at the National Gallery, London on September 14, 2009. The participants at that meeting were invited to consider the Foundation’s efforts to identify ways in which several “pilot projects” that had been launched under the auspices of MAC, and were poised to develop into “prototypes,” might become both sustainable and cost effective over the long term. 

The BM was represented at that meeting by Dominic Oldman, Deputy Head of Information Systems, and principal investigator on the present proposal.  Immediate synergies were recognized between the shared infrastructure concept and work already underway at the BM, which is actively working to fulfil its public mandate to make its collections data broadly accessible via the internet.  As a result of Oldman’s inclusion in these early discussions, the ResearchSpace concept emerged, and between September 2009 and the present has undergone a period of informal review and refinement. The BM, with Oldman’s leadership, now proposes to move ahead with a formal planning and implementation process to bring ResearchSpace to fruition.

Project Rationale

A shared infrastructure is seen by the Mellon Foundation as essential to the sustainability of the MAC program prototypes because it would control escalating software development as well as (especially) maintenance and operational costs. Due to the nature of these pilot projects as experiments in new methodologies, it is understandable that each team has selected a different approach, and that they develop their required technology independently from the others. However, analysis and review have shown that the technical requirements of these projects are in fact, quite similar, though they are implemented using different platforms and back-end databases; therefore, a move toward a single technology platform would allow for a dramatic increase in sharing and collaborative development of new components.  It would also reduce the need for institutions to fundraise independently to procure and support a wide range of different technology solutions.

The shared approach that would be provided by the ResearchSpace environment would reduce both time and costs (for example, the time and costs involved in setting up a separate IT infrastructure for each project and savings made by reuse of existing technology) as well as diminish the development risks for any particular project. The gradual building up of these ‘community’ tools and services would allow the prototype organisations to concentrate on the collection, organisation, and presentation of their data, rather than deal with technology decisions and issues that most of them are not equipped to manage. ResearchSpace would also be broadly available to any cultural heritage organisation, regardless of their funding source, facilitating the creation and sustaining of a large community of people, data and research tools. ResearchSpace would be owned and managed by the organisations that use it rather than being managed by any single institution.

Oldman and the ResearchSpace collaborators have recently become familiar with Project Bamboo (www.projectbamboo.org), which received support from the Mellon Foundation in 2008. Bamboo has similar objectives to ResearchSpace in that it seeks to bring together collaborative communities using shared services, though its focus is primarily on the Higher Education sector as opposed to museums and other cultural heritage entities.  The scope of Bamboo currently seems wider than that envisaged by ResearchSpace, and as such, its approach is significantly different in terms of the resources, framework, and technology required. A meeting was held between the ResearchSpace and Bamboo project teams to understand the potential overlaps and differences between the two projects with a view to informing both projects of possible synergies, either now or the future.  The meeting report is included at Appendix H.

Project Description

The planning tasks proposed for this Officer’s grant depend in large part upon active collaboration with the project teams currently funded by MAC:
•    The Courtauld Institute of Art: Master of the Fogg Pieta / Maestro di Figline Project
•    The National Gallery, London: Raphael Research Resource
•    The Rijksbureau voor Kunsthistorische Documentatie (RKD) with The Royal Picture Gallery, Mauristhuis: The Rembrandt Database
•    The Stiftung Museum Kunst Palace: Cranach Digital Archive

Taken together, these projects represent a coordinated effort to create new digital resources that are able to interface with those of other institutions and involve the gathering, storage and dissemination of cultural and heritage content.
The following work has been completed to date, largely by Oldman, and in collaboration with the above mentioned partners:

  1. An initial Project initiation Document was written on 26th November 2009 describing the first stage of the project and serving as the basis on which to define the ResearchSpace specification and provide on-going recommendations. (See Appendix A).  This document is not current, and the plan not acted upon, because of organisational restructuring within the Mellon Foundation. However, it contains useful information which has been reviewed by the project team and is valuable for future reference. Some of the objectives in this document are consistent with it.
  2. Two further documents were drafted (both attached), one describing the conceptual framework for ResearchSpace and the other describing how it would support scholarly research (see Appendices B & C). These descriptions are current and remain the foundation for future objectives and planning.
  3. A design meeting was held in London on December 9-10, 2009 in order to gather the initial requirements of stakeholders, primarily from the Mellon prototype projects.
  4. An initial Requirements Catalogue was compiled by consultant Sam Wood, a draft of which is currently being circulated among the principals of the Mellon prototypes for revision, refinement, and prioritization of the core requirements.  A final revision is expected at the end of April 2010. The current version is attached at Appendix D.
  5. Presentations have been made to the Directors of the British Museum, Rijksbureau voor Kunsthistorische Documentatie (RKD) and The Royal Picture Gallery, Mauritshuis.
  6. Meetings have been held with the Director of the Courtauld Institute of Art and the Director of Information at the Tate.
  7. The BM’s internal pilot project testing the applicability of using RDF (semantic) data to a defined set of museum data was completed at the end of March 2010.  The results of the pilot are now live on the Museum’s Collection Online system.

Priorities for Project Preparation

This proposal provides the opportunity to evaluate and clarify aspects of the ResearchSpace environment in advance of full project execution. As such it attempts to identify and prioritise the work still necessary to support a fuller picture of the final project plan, and provide mitigation of the main risks. There are three main elements to the ResearchSpace project. The following provides a summary of each element indicating their relative priority for preparatory analysis. 

Collaboration Environment - Collaboration software is now well established in both open source and commercial communities. In addition, development against these systems is proven (for example CollectionSpace uses the ‘Nuxeo’ content management system). The project should therefore be confident that a robust evaluation process of leading content management systems that support collaboration services will be sufficient. 

Research Tools – It is anticipated that many different research tools would be developed for the ResearchSpace environment over time.  Each individual tool would carry its own level of development complexity and its success would depend on the robustness of analysis and design applied to it. Initially the ResearchSpace project should easily be able to provide a basic range of useful and productive tools leaving others to be developed gradually. The type of development work is considered to be more standard and reflect established functionality already proven in other environments, for example, the Zotero plug-in system.

Data Services - RDF technology has been available for some time but the significant increase in its use has been relatively recent. Some RDF tools are well developed but overall development of the technology continues. Although institutions using RDF are now not seen as early adopters, its use is still considered innovative. The British Museum has mitigated its own risk by conducting a pilot within Collections Online that is now live on the internet.

Given that the more standard elements of ResearchSpace development are reliant on interfacing with an RDF data store, further end to end design and feasibility work of this technology, based on a larger set of data to inform the main project, would be prudent. The RDF element is therefore considered by the proposal as a relatively higher priority for preparatory work because it will provide important information for assessing feasibility, risk and cost. As such the technical work in this proposal deals with data services and other components would be developed more fully in another later phase of the project.  

Objectives

The present request would support the preliminary work needed to define and specify the resources required to initiate and complete full implementation. The objectives are as follows:

To specify a shared hosting model (and costs) and determine a shortlist of appropriate hosting providers who would be willing and able to support the shared technology model. The hosting model would require the agreement of MAC project stakeholders and the Museum would use its own hosting environment as a benchmark for selection.

  1. To determine appropriate criteria and evaluate and choose a RDF triple store management system.
  2. To provide a specification for the shared technology environment. The specification should include data and business process, software development, and collaboration and workflow and should be costed. The Requirements Catalogue mentioned above informs some elements of the specification.
  3. The specification would include, if appropriate, existing open source components for use in the project, subject to evaluation. For example, various open source collaboration suites (Sakai, Alfresco) and other related components (like Zotero) may already meet some ResearchSpace requirements.
  4. To specify and, if possible, develop and test a working data publishing prototype to allow cultural and heritage organisations to publish their local collection data to a common schema that can be converted for use within the ResearchSpace environment.
  5. To specify the processes and model under which collection information would be stored, manipulated and published in RDF format. 
  6. To investigate different infrastructure options and their likely ongoing cost (comparing them with current MAC project costs) thereby indicating the level of sustainability that can be achieved.
These objectives will be taken as far as possible with the level of resource and budget available and as such regular updates on progress and actions will be provided. Objectives 5 and 6 will be expanded to provide a detailed brief to the Seme4 consultancy which will be attached to the contract for services. CV’s will be obtained for Seme4 staff working on the brief which is expected to include the delivery of the following elements; 

•    Priority 1 - The provision of a XML/RDF schema designed to support the process and movement of data between participating institution’s local database systems and ResearchSpace. The schema should be based on the British Museum’s data schema and support the implementation of the CIDOC-CRM ontology framework.
•    Priority 1 - A detailed and annotated architecture diagram (with technical documentation) showing how data in ResearchSpace would be edited, annotated, published and generally managed (with authority, version control and security).
•    Priority 2 - A prototype system implementing the process above using the British Museum’s data.
•    Priority 2 - Performance testing of the RDF (using the full set of BM collection records) benchmarked against traditional systems (like the Collection Online relational database) using appropriate indexing.
•    Priority 3 - If possible, some prototypes implementing some of the data management processes above.

Timetable



Budget Narrative

The requested Officer’s grant would be used primarily to support consultancy fees for the following specialists:

•    RDF and OWL (Web Ontology Language) specialists
•    Business Analyst (IT bias)

It is expected that objectives 5 and 6 would be completed using consultancy from Seme4 Ltd (a consultancy set up by Southampton University staff). Other objectives would be achieved through the contract employment of an independent Business Analyst  (most likely employed through an IT recruitment agency) who would deliver the main specification documents and recommendations that would be used to plan and specify the full project. This type of post is relatively standard and the author does not anticipate any difficulties in recruiting in the current job market.

In relation to the quote from Seme4 Ltd, a single tender process has been adopted and approved by the Museum’s Director of Administration under Museum tendering rules. This procedure is allowed in the following circumstances:

  1. An innovative product or service is offered by the supplier
  2. An overriding need to maintain continuity or the introduction of another supplier would cause excessive cost or disturbance to the operation.
It is currently unknown whether all of these deliverables can be met by the current timetable and budget. As such the deliverables are prioritised and will be finalised when negotiating the contract with Seme4 Ltd. The Seme4 quote for the work is provided at Appendix F.
The British Museum intends to provide some support for further analysis and manipulation of its collection data. This would involve the following resources:

•    Programming analysis – from the existing IS programming team
•    Collection data expertise – from the Museum’s collection specialists.

The project will be submitted to the British Museum’s project gateway process and be reviewed by the Museum’s Project Board chaired by Andrew Burnett, Deputy Director. Any intellectual property produced with support from the Mellon Foundation as a result of the project would adhere to the Foundation’s intellectual property policy.


Appendix A – Original Shared Infrastructure PID – 26 Nov 2009

Contained in separate PDF document


Appendix B – ResearchSpace Conceptual Framework – 3 Feb 2010

Contained in separate PDF document


Appendix C – ResearchSpace Research Services – 12 Feb 2010

Contained in separate PDF document


Appendix D – Draft Requirements Catalogue

Contained in separate PDF document


Appendix E – Business Analyst (IT) – Draft Summary Job Description

Contained in separate PDF document


Appendix F – Quote from Seme4

Contained in separate PDF document


Appendix G – British Museum

Contained in separate PDF document


Appendix H – Bamboo Meeting Report


Comments