The goals have two tracks:

1. Migrate existing NORA D2K itineraries to SEASR so that the existing NORA-OL would remains a viable application: This would be a good side affect of the trying out new SEASR infrastructure, the resources have to be spent to integrate it with the MiddleWare/Proxy layer thus it will be available to NORA-OL as well.

1. Create a kick ass demo that uses MONK components inside SEASR environment and pull those components into SEASR as standard humanities based components.

Priority: [To be determine with SEASR team]

Demo Ideas

-attributed to Loretta:

Setup a production SEASR infrastructure

-Auto Builds
-A production install with simple itineraries and where the NB/SVM itineraries can go

Implement Naive Bayes and SVM flows

-Use Weka components for both of these flows.

Entity Extraction components

-Use Open-nlp for creating such a component.

Wrap the collection/chunk selection and browse components from MONK into SEASR

-This step involves investigating web application as a part of the SEASR component and MONK workbench as a part of such
an experiment.

FEDORA integration with MONK components via -Middleware/Proxy

• Ask Mary to update information here.

SEASR DEMO (text/xml)
SEASR DEMO (text/xml)
SEASR DEMO (text/xml)