Changes between Version 1 and Version 2 of Discovery

02/10/08 16:25:39 (13 years ago)



  • Discovery

    v1 v2  
    1 The [ NDG discovery portal] provides a web portal accessible by standards compliant browsers that allows a user to freely discover datasets held by data providers who have offered up appropriately formatted discovery metadata for harvest by the discovery portal backend.  The current discovery architecture is outlined here: 
     1= Discovery = 
    3 [[Image(source:TI01-discovery/trunk/doc/NERC Metadata Gateway.JPG)]] 
     3(See the OldDiscovery page for NDG2 era information.) 
    5 Where NDG compliant data providers hold datasets, the portal will provide information about NDG services available for those datasets, and provide links to instantiate these services. 
     5The main aim of the NDG3 discovery activity is to improve the functionality of the NDG discovery service (and fix any bugs involved in improving the speed and reliability of the discovery service which occurred under the auspices of the NERC portals project funding). 
    7 The key NDG code in the discovery pipeline is: 
     7=== Meetings === 
    9  1. The OAI ingest [source:TI01-discovery/trunk/ingestAutomation/OAIBatch code]. 
    10  1. The [source:TI07-MOLES/trunk/JavaCode/ code] to go to miniMoles. 
    11  1. The [??? code] to go from mini-Moles to DIF. 
     9==== October 01, 2008 ==== 
     11The discovery work will be done by Steve D.  
     13Expect 60 days work between now and the end of March. To be confirmed. (Matt/Steve?).  
     15The following activities are in scope, but prioritisation and job quantification has yet to occur (it is unlikely that all will be doable). 
     17 * Implementing spatial proximity for bounding box searches as the default. 
     18 * client support for all the options provided by the server. 
     19 * introducing logging:  
     20  * all searches to be properly logged. 
     21  * all outbound links to be modified so that they are logged before redirection. 
     22    * consider whether this information could be used for search ranking (this would involve some code to parse the logs, and update the dataset entries some sort of link hit count, presumably daily. Would need to ensure robots were ignored). 
     23 * Improve vocabulary server calls for multiple parameters.