Thursday 24 February 2011

Pelagios Project Plan Part 3: Risk Analysis and Success Plan

Risk Analysis

The main risks identified are shown below and will be reviewed every month:

Risk

Prob.

Impact

Action to Mitigate

Overly-complex or simplistic ontology inhibits uptake

Low

Med

The domain and technical expertise of LUCERO, ACRG and Pleiades, plus concrete use-cases from each partner, ensures a robust ontology. The use of modular extensions means that problems specific to one document type will not affect others, permitting multiple interest groups.

Lack of community interest

Low

Med

Each project partner already comes to the table with an established community of users, among whom the Pelagios outputs will be rolled out. Other interest groups such as Google, Ordnance Survey, GeoNames, EDINA and the European Commission will be involved through participating at the workshop.

Project stream unable to fund mapping and publication of data

Med

Low

The global economic crisis threatens all academic funding, but the highly international nature of the consortium and low dependency across the Work Packages greatly spreads this risk.

Critical failure of Pleiades Project

Low

Med

Pleiades is a core element of the Institute for the Study of the Ancient World (ISAW)’s digital infrastructure with high level institutional support and long term funding. It will be possible to use URIs generated from the Princeton Encyclopaedia of Classical Sites as a fallback position if necessary.


Evaluation

The success of the core ontology for place references (COPR) and linked open geo-data (LOG) approach will be measured against each publication stream using SMART criteria:

Specific: The ability to map a given document type to i) persistent Pleiades URIs, and ii) COPR-compliant RDF (see project plan post 1)

Measurable: A comparison of the precision and recall of the linked open geo-data (LOG) against a benchmark manual mapping of a sample from the source material

Achievable: Each team will identify an appropriate corpus in terms of both complexity and volume to demonstrate the value of the ontology

Relevant: Publication streams will refer to ancient places centred on the Mediterranean (whence come the majority of ancient sources) and datasets likely to be of interest to a wide audience

Timely: Sample data will be available four months prior to the end of the project. Complete datasets will be available one month prior to the end of the project.


In addition, documenting the process of implementing the core ontology by each stream is a key output of WP2 (see project plan post 6) by which future requirements can be identified, while each partner group will trial and elicit feedback on the demonstration web services among their target user-groups.


Sustainability

The RDF ontology will be hosted by Pleiades, whose long-term strategy and funding ensures the stability of both the domain name and content. Instance data will be held by the relevant partners, each of whom have sustainability at the core of their funding and institutional commitments. All web services and tools will use Open Source technology and Open Standards so as to be freely available for adoption by third parties. Code will be hosted in a suitable repository such as SourceForge or GitHub. A demonstrator version of the software will be hosted on multiple websites maintained by the project partners.

No comments:

Post a Comment