31.07.2012 WP10 - X-Search Integration

From D4Science Wiki
Jump to: navigation, search

Agenda

Time: Tue, July 31, 2012, 15:00 - 15:30 CET

  • Progress of integration activity between X-Search and gCube Search
  • Planning of remaining activities in accordance with time constraints

Participants

  • Alex Antoniadis (NKUA)
  • Gerasimos Farantatos (NKUA)
  • Rena Tsantouli (NKUA)
  • Giota Koltsida (NKUA)
  • Pavlos Fafalios (FORTH)
  • Yannis Marketakis (FORTH)
  • Yannis Kitsos (FORTH)

Discussion Summary

Integration status

  • X-Search portlet
    • Activity almost complete
    • Result retrieval from gCube Search
    • Communication with X-Search Service by means of gRS2
    • Result presentation
  • X-Search service
    • HTTP RESTful service
    • Development instance currently hosted at FORTH
    • Performs functionality of result clustering and entity mining

Remaining activities

  • Result presentation and the functionality of the X-Search portlet in general must achieve the standards required for usage in the production environment and especially for demonstration during the review.
    • Achievement of this goal will be given high priority among the activities during this week. Specifically for the presentation of content to the user, NKUA will support FORTH in the usage of Content Viewer by the X-Search portlet.
  • The provided functionality currently exploits result snippets. Result analysis over all content using annotated search fields is not supported yet.
    • A discussion regarding the effort needed to include the feature in the final components followed. NKUA felt that this activity is feasible as the effort required is not significant, taking into account that all supporting functionality is available through the Resource Registry.
    • Both partners agreed that the feature should be included in the activity. In order to facilitate integration, NKUA will provide support on the usage of Resource Registry.
  • Integration with FLOD SPARQL endpoint depends on a ticket managed by FAO
  • Discovery and registration of the final version of the X-Search Service to be released should be discussed with CERN.

Further technical topics included

  • Search metadata handling and presentation in the context of result presentation: Discussion included the field model exploited by the Search System, the correspondence of fields with the fields exposed by collections and the functionality of presentation fields.
  • The possibility of performing tests in the production environment. This could enable us to perform tests with the data hosted in production and take advantage of a more stable environment. Most issues encountered during last week in the development infrastructure concerned the failure of Search to answer some queries, some downtime of the Search Service and some problems related to the Information System. However since the first two issues were resolved last week and the third was a temporary problem which was fixed quickly by CNR, the development infrastructure will be able to support the extensive testing that will take place in the coming days. Regarding the exploitation of the production environment, configuration issues were discussed especially for Resource Registry and both partners agreed that this possibility should first be discussed with CERN.

Actions

  • In the coming two days FORTH will work on result presentation in order to attain the standards required for demonstration of the functionality, e.g. functionality and look and feel compatible with the current results portlet
  • In parallel, FORTH will be communicating with NKUA in order to receive support regarding the usage of Content Viewer and the integration with Resource Registry.
  • FORTH will communicate with CERN to discuss details regarding the publication and discovery of the X-Search Service in the production environment and work on any suggestions provided by the latter.
  • Progress and the actions to be performed during the last week of August will be discussed this Friday.