Biomedical Engineering Reference
In-Depth Information
My Publications. In an organisation the size of AstraZeneca, it is
diffi cult to track what experience the employees have, especially when
that experience was gained before the employee joined the company.
For scientists, one way of tackling this problem is to have a view of
what publications these people have authored during their career.
However, even viewing this is diffi cult as, very often, the author's name is
not unique (e.g. try searching for 'N. Brown'). One of the fi rst searches
most scientists perform when they come across the SOLR Publications
Management Index, is for their own publications. A small system was
therefore developed known as My Publications that allows users to
quickly perform this search and then to tag the documents with their
unique AZ identifi er. It is then possible for other people within AZ to
search for publications that are authored by AZ employees and to
discover their areas of expertise based on the areas in which they have
published.
This concept is to be extended to include internal documents that allow
users to quickly identify other people within the company who have been
(or are) working in a particular area in which they are interested. It is
hoped that this will encourage a culture of interaction and collaboration
that helps stimulate innovation.
14.9 Challenges and future developments
The SOLR Publications Management system has proved to be a robust
and reliable system. However, if it is to be made available to the wider
AstraZeneca community, the following enhancements need to be developed.
￿ ￿ ￿ ￿ ￿
The system should be made more robust. It may be possible to use the
new SOLR Cloud development to provide this [10]. This will allow
data replication across multiple servers, and failover facilities. In
addition, it is desirable to implement separate index and search servers,
to improve the amount of data that can be indexed, and the frequency
at which the index can be refreshed. This is especially true if internal
content is integrated with up to 200+ million documents.
A system management facility should be implemented. Our current
systems check to see if the servers are still running, and re-start them if
necessary. However, this is still a mainly manual process, and the
search for third-party tools to support this (e.g. LucidWorks [11]) is
desirable.
 
Search WWH ::




Custom Search