Data Source Owner's Discussion

Meeting: 16 November 2012 - Ochil Room

Attendees: James Toon (PURE), Steven Ross (CAM), Ruth McCallum (Academic Registry), Neil Allison (UWS) and Craig Middlemass (Project Services)

Presentation Slides: Data Source Owners Discussion

Main Discussion Points:

• Filtering and faceting would be very useful in allowing discovery of content, allowing the search to be refined without having to enter a new search term.

• NEED TO CONSIDER - What facets to use? A single facet per data source is not conducive to navigation, would be better to target audience of broader Theme or Subject.  For example an “Applicant” facet could target SSFS, Accommodation and Studying content.

• Having a context sensitive search would be useful to assist in getting to the right information, although context sensitive search study (conducted by Neil) revealed that 55% of users wanted to broaden their search and 45% wanted to go deeper so it is not always true that users want to narrow down their search within an area of the site, although it was not ruled out to present a facet for the context sensitive search but default to having the main UWS facet being in focus.

• NEED TO CONSIDER - Mapping university terminology to real life terms.  For example “matriculation” is a very specific term how would we say this in a non-university context – “Do you mean ‘matriculation’?” means little to people outside the University.

• Push of events and press releases via widget (side links on search to promote University related content) – needs to be non-intrusive to main search body.

• Need to retain a University look and feel, pushing information using spotlights and widgets have a very commercial (sell, sell, sell) feel about them; need to be careful that they are used appropriately and don’t become the mainstay for promoting content.

• NEED TO INVESTIGATE - Target Audience, how do they end up at the University site or sub sites? The flow of navigation, how do people get into the site, context sensitive searches are useful for the sheer volume of content available

• Need to ensure that where we want to push a type of content that we offer the benefits of adopting a simple schema / metadata format that will allow all sources of events/new/etc to be pulled together in view within the search and it is not cumbersome to create this type of content

• CONSIDER - Geographical filtering to find things close to you or in certain locations – Southampton has done something similar with street maps

Summary of Systems Discussed

PURE

• Research explorer for PURE is targeted toward discovery of content rather than searching.

• PURE contains lots of tags (metadata) but suffers from limited filtering and no faceting of search results.

• Drilling down through the search is mainly used to find the content for download or viewing (this requires heavy use of tagging to present the most appropriate results).

• Linking strategies are used in the site to assist in navigation to useful / targeted content.

• Download statistics from Pure are used in addition to Analytics to target content and promote.

• Full text indexing is available but not heavily used.

• Audience – Public and Academic (Global Community)

• PURE has funding to further enhance their search and discovery features

• Like the look and feel of www.freelibrary.org

CAM

Publishes CMS driven content source systems except:

• Online Prospectus (difficulty with searches due to way that information is published from MySQL database via PHP and proxied into Polopoly – this loses the metadata on route meaning that all meta content effective looks the same to a search engine)

• UG Prospectus has custom search that uses keywords and title queries for programme discovery.

• PG Prospectus does not have custom keyword search. Reliant on standard search functionality.

• Events – Google Calendar – The information may not be index-able (searchable)

• Campus maps – Exposure of building information and locations within search results

• Video content – provision for video discovery needs to included

ACADEMIC REGISTRY SYSTEMS

• DRPS – how are student directed to find content? Visiting students via application links; Links though from KIS to Degree Finder and then from Degree Finder onwards.  Degree Finder links through to DPT / DPS information for more granular information on what courses and programmes are available to study

• DRPS used by academics to ensure regulations

• DRPS used by students to drill through information on specific courses or combination of course (terminology minefield “module” and “course” are often interchanged between institutions).

• NEED TO INVESTIGATE - Do we have any analytics on search of DRPS information?

• A project is being planned for 2013/14 to revisit the DRPS and what is published and how – may want to interact with this project to ensure metadata/schemas are considered as part of the publishing.

• SSFS Scholarship database is an access database that publishes information using cold fusion - may need to consider uprating infrastructure if wanting to hook up the database rather than crawling the web pages.