DB interface meeting nov 1st 2011

DataBank Interface meeting 1/11/11

Attendees:

Neil Caithness Katherine Fletcher Neil Jeffries Anusha Ranganathan Joseph Talbot (chair)

Introduction

The purpose of this meeting was to discuss the interface to DataBank and how we might improve it. We started by discussing the purpose of the system and then began to discuss the system in terms of what audiences it needs to support, what aims each audience would have how these audiences could be grouped and prioritised.

Audiences

Very important

Audiences it is critical that we support.

Academic users

• Researchers (as data owners)
• Researchers (as browsers of the system)
• Reader of a paper

Academic Administrative

• Research services
• Departmental administrators
• Academic Supervisors
• Funding Councils

automated interactions

• robot harvesters
• harvester designers

other users

• prospective users
• librarians

Quite important

Audiences which we should consider

• Curious members of the public
• School teachers
• University research commercial development units, e.g. ISIS
• Biographers (of researchers)
• Media

unimportant

Audiences which we recognise, but don't need to make special consideration for.

• Research sceptic
• Lawyers (dispute resolution)
• Patent agent
• Police

conclusions

An analysis of our audiences provided the following goals for our interface:

• provide a secure repository for research data
• support direct linking to datasets by external systems (e.g. a publication)
• support the harvesting of metadata by external systems to maximise discoverability of datasets
• provide an API to allow external systems to provide bespoke interfaces to specific datasets
• provide a users interface for searching and browsing datasets
• communicate the system's purpose, features, strengths and limitations

In order to fully achieve these goals, the DataBank interface needs to be improved in the following ways:

1. Make the general interface more intuitive and usable to the general user. This can be achieved by: a. improved navigation and general page design b. adopt a more concise, plain English writing style

2. Improves the searching and browsing of datasets. This is the tricky one, as the system doesn't mandate much metadata, we can't provide much in the way of structured searching or browsing. There are some things we can do though:

   a. be transparent about the fact that searching is only as good as the metadata and that quantity and quality of metadata will vary between datasets
   
   b. provide a list of keywords within the system (at the silo and package level),  collected from a keyword field (note: this impacts on DataStage - we will need to add a keywords field to package metadata) and harvested from the description field.
   
   c. support searching datasets for a person, i.e. from an ownership/authorship angle.
   

3. provide better explanation of the system's purpose, features, strengths and limitations (on the home page and within context of particular features.)

4. ensure that all present and additional interface features are as accessible as possible.

additional features and interfaces

1. keywords interface a. list of keywords b. ability to search/filter list of keywords c. browse/search datasets based on keyword(s)

2. search interface a. search page which outlines search options i. Person search - only searches fields relevant to ownership and authorship ii. search silos, packages, filenames (checkboxes? Radio buttons?) b. results page - what to display on search results page

Actions

Joseph Talbot (JT) Anusha Ranganathan (AR) to meet to discuss how best for JT to work on interface.

AR to show JT templating system. Possible areas to cover:

* making site-wide template changes (e.g. changes to header, footer, etc.)
* creating new pages
* adding values into a page (e.g. adding owner in a list of files)
* creating links within the system

AR to add ability to include keywords from keywords field and harvest keywords from description fields

AR to add keywords interface (list of keywords linked to search results for that keyword, ability to filter list of keywords)

AR to build search page

JT to design improved page layout and navigation

JT to work on improved interface design for:

	1. homepage
	2. browse silos
	3. silo page
	4. package page
	5. browse file structure

JT to consult Dev-list re text for describing DataBank system.