Primo - product update

[live-blog IGeLU Conference 2013, Berlin - please forgive typos etc.]

Gilal Gal, Director of Product Management, Ex Libris

In Primo, we focus first on the "engine" to make it functional. It is pure engineering, ensuring movement in the production. With Primo Central we need to ensure it works 24hrs a day, it is quite an achievement but takes a lot of focus. We are the software vendor, we don't sell content, but we can't talk about Primo without talking about the content because we have to handle it correctly. Primo is the tool through which you give services to your customers and we take into consideration comments from users and yourselves, as administrators.

Breath of development in Primo:
We develop OPAC-type functions, such as browsing. We rely on bx, which is contextualisation of searching based on preferences. We listen to the community's requirements, such as a FRBR presentation, added functionalities such as direct request for photocpies/digitisation etc., ability to update the password. We continue to invest in the mobile interface.
We were asked to invest more in the administrator service (i.e. not via the command line) to load files etc. so we have improved those services. We will also continue to invest in security, we've  achieved ISO 27001 certification for informaiton security management. We are using Big data infrastructure (Cassandra and Hadoop) for data processing. We will move to the multiple SaaS infrastructure in a future release. We are not abandoning Oracle but for more sophisticated things, some of the data structure will move to Cassandra.
We have a Primo Central resource activation to facilitate management of resources. We will continue to improve the performance of specific things, such as getting results faster so that the search experience for the users is improved. One of the things we want to develop is the browse virtual shelf functionality.

In terms of content we are adding abstract  indexing collections. We are including published research because that's what people want more, but we are also including things such as thesis, technical reports (unpublished scientific research), raw material for research (Mintel, Data-Planet) etc. Primo Central enhancement with phantom collections - activate collections via alternative coverage.

Access to the Primo Back Office is made easier. It is done in 3 easy steps, getting an email. information through a webform and receive your completion email.

There is also syndetics-type information for a search carried out. This is presented as a pop-up box on the side of the screen, providing basic information based on the search from reference material, which includes relevant links to other relevant sources of information.

Tamar Sadeh, Director of Marketing
Scholar Rank

See presentation given last year. We have done some evaluation and have used qualitative and quantitative methods.We looked at useage, KPI based on session/GetIt (=concrete interest in a specific item), location of the selected record, time until selection, use of facets and navigation to the next page etc.  Globally we see that there is an improvement in all areas. Personalised ranking can be on or off.

Open Access is another topic giving new opportunities.  Publishers take part in this movement especially in Gold Open Access. The proceedings of publishing are the same but the money comes from the researchers, not the reader. The Green model is valuable for institutional repositories. What we do in Primo is that we highlight what's open access. From the point of view of readers, it doesn't matter where an article comes from . The interesting thing is with hybrid journals. Those are based on subscriptions but some articles are free. We encourage institutions to promote their work via Primo. It exposes the content but also links to other relevant work (same other, same subject etc.)

Question and Answers

Q: Will there be a more elaborate way of depositing institutional repositories , i.e. more information than just author, title etc.?
A: We aren't limited to any kind of information and we would be happy to extend this to anything you need if you tell us about it.

Q: Installing Hadoop and Cassandra for SaaS customers, what about those that have local installations? How will you ensure that the performance will be acceptable to both groups?
A: It should be seamless to our customers, the software will identify if it should use Oracle as it does today or if it should use Cassandra/Hadoop

Q: Phantom collection: will we find the name of the collection in the facets?
A: To create a facet we need to indicate which collection it belongs to and in this case it would be collectionS so I'm  not saying no to the question but we need to work on it

Q: Primo Central and Local are separate indexes that can't be merged. When there is a blended search, it is not possible to dedpuplicate and RFBRise across both indexes. Also ScholarRank and Personalised Ranking only apply to Primo Central. Are there any plans to imrpove integration of both indexes?
A: Yes we have plans to have ScholarRank in Primo Local. We are looking into the FRBRisation. Until the collection wasn't that big but this is changing and we will be looking into this.

Q: The subjects and author names are copied "as is" from data sources into the index, this makes itimpossible to search on specific subjects or variations of names. Are there any plans to normalise this (authorities etc.)
A: We put a lot of effort in normalisation but it is difficult to be perfect. We monitor our work in this area, we are starting a big scale pilot of normalising this data, using a software infrastructure dedicated to that. We did a small pilot and are now going to start a big pilot, also working with publishers.

Q: A Worldcat adapter is available to order to search directly from the Primo UI. This could be very useful, if the description was based on records with more information than a slim DC and if facets were created. As it is now, it's hardly interesting for our users.
A: Difference between the API and the search engine. Worldcat don't provide facets and answer the query in rank order. We can't provide a facets option based on that. We have the capability but we need to review how we will do that.