IR Metadata Teleconference Thursday, 18 April 2013

Date: Friday 18 September 2009

Time: 2:30-3:30pm AEST (NT & SA 2pm; WA 12:30pm)

Participants: Julie Woodland (Curtin); Joanne Croucher (UNSW); Andrew Harrison (Monash); Sten Christensen (USYD); Danny Kingsley (ANU); Patricia Scott (Deakin); Fiona Burton (Macquarie); Natasha Simons (NLA); Basil Dewhurst (NLA); Ann Huthwaite (QUT); Margaret Henty (ANDS); Peter Sefton (CAIRSS); Tim McCallum (CAIRSS); Kate Watson (CAIRSS).

Apologies: Rebecca Parker (SWIN); Simon McMillan (UNE); Vicki Picasso (NEWC).

Schemas in use in Aus IRs

- Based on the data collected by CAIRSS from the phone discussions with each repository manager, most repositories are using DC and MARCXML, with 6 IRs using MODS http://www.loc.gov/standards/mods/.

- Kate (CAIRSS) commented there were some sites using MODS from the start, but also appeared to be a trend emerging as with sites looking to move across to MODS for enhanced capability.

- Andrew (Monash) discussed his recent experience in moving to MODS for the Monash IR. With the various object types in their IR found DC not rich enough for all special collections, and wanted to utilise a richer schema. Andrew discussed the MODS extension capability (to feature local and other elements). Patricia (Deakin) spoke about their use of MODS and this extension capability (utilising this for both a copyright md and deakin md extension). Andrew discussed batch loading data from Research Master into IR in MODS. Andrew commented on finding MODS granular enough to match-up with the Research Master data. Pat commented Library of Congress website provides good MODS support documentation. Joanne (UNSW) discussed how MODS is utilised in other UNSW repositories (not yet in research repository).

Q1. Any other repository managers thinking of using MODS in the future? Any information on your experience/research into moving to MODS? Community interested in hearing about it.

MACAR past use/present use

- Discussion on MACAR http://macar.wikidot.com/ resources use (past use and present use).

- Katy (CAIRSS) shared feedback from her discussions with CAIRSS repository managers on MACAR resources use (see notes in table below):

Some have not used MACAR resources in the past. Others have used MACAR work for mappings.

Some discussed how consistent metadata important for sharing data in the future, while others commented total standardisation is not realistic across the AUS IR community.

Some discussed how institutional requirements will override MACAR guideline integration.

Some found MACAR advice did not work well for creative works resource types for ERA.

Many commented advice needs to be realistic and practical, while others noted MACAR tendency of purist viewpoint not always able to be applied in a practical IR situation.

Some commented would like to see MACAR continue in some form, while others noted while a need for metadata discussion of some sort not sure if a permanent group of people the way to go. Some notes the need to get the mix of people right for advice to be relevant and practical.

Some commented they would prefer a group to highlight options rather than provide a framework.

Many commented the advice needs to change to suit the changing environment.

 

- Discussion on MACAR resource types not being extensive enough. Some felt they needed a finer level of delineation (e.g. thesis), while others noted MACAR had left a broad term intentionally so institutions could narrow down further using other metadata fields (not resource type more genre field). Some commented this was difficult to do with certain software which needed sets for harvesting to be delineated by resource type field.

- Discussion on perceptions of standardisation happening in IR metadata in the way it has happened with traditional library cataloguing. Some felt it would not be possible.

- Discussion on how repository managers will often go with the MACAR resource types where they can, but institutional and software requirements will override how local resource types are set up. Constraints in the ability to go with the purist metadata.

- Discussion on DC type only being able to have 1 attribute.

- Discussion on the difference between local metadata needs and global metadata needs. On how institutions need local metadata as well as global metadata standardised for harvesting. Discussion around how current IR software limitations in delivering this.

- Discussion on the need to separate the software limitations from how you would actually like the metadata to work, to help move beyond those limitations.

- Discussion on whether there could be a loose agreement by Aus repository managers to (1) adhere to DC types; (2) make best effort (taking local metadata requirements in mind) to adhere to MACAR resource types. People discussed that (1) was probably more realistic. Could there be a loose agreement on base level of resource types to use?

- Discussion how for ERA resource types need to be at granular level, whereas for harvesting resource types need to be at global/common/simple level for maximum exposure. Andrew (Monash) shared example of how at Monash they are using MODS for local metadata and DC for global/harvesting metadata.

- Discussion on if it would be better to stick to International broad descriptions rather focus on an Australian MACAR set.

Q2. Do MACAR resource types need to be extended?

Q3. Could there be a loose agreement by Aus repository managers to adhere to DC types?

Q4. Could there be a loose agreement by Aus repository managers to make best effort (taking local metadata requirements in mind) to adhere to MACAR resource types?

Q5. The need to outline how we want the metadata to work, regardless of current software limitations – to help move beyond those limitations in the future. Your thoughts/comments?

Q6. Is there a need for a MACAR group be continued in 2009/2010 (with facilitation via CAIRSS)?

Resource types and NLA normalisation

- Discussion on the normalised view of Australian Research Online http://research.nla.gov.au/ versus the actual view of all the different resource types harvested by NLA.

- Explanation of how the NLA normalises all the data (including resource type) harvested and how the final end view is not what is originally harvested there is editing performed by the NLA (normalisation) for this view to appear as such (and therefore maintenance and cost). Basil (NLA) provided some background information on how the metadata harvester was built internally and harvests OAI records with a number of processing steps. These applied transformations allow the NLA to have power over how the data harvested is displayed. This process utilises the MACAR resource types and turns quite obscure types into more common types. NLA performs this to improve the discovery of resources in IRs.

- Discussion on how resource types in IRs vary (those that follow the MACAR resource types and those that do not). Constant review for NLA as IRs add new resource types outside DC and MACAR.

- Peter (CAIRSS) mentioned it would be good for repository managers to have an awareness of this normalisation happening at the NLA end. Peter thinks it would be useful for repository managers to view this harvested data without normalisation to generate further discussion on this topic. The CAIRSS information on SNAPSHOTS outlines a possible method. Peter discussed how this normalisation should ideally not be happening at NLA idea that repository managers should be able to control it with their local and global metadata view long term (although some repository software not allowing this at present).

- Discussion on statistics on how many end-users use the search filter resource type. Discussion on if normalisation is worth the effort? Natasha (NLA) advised resource type was not the only normalisation they perform on the harvested data.

Q7. Your thoughts on the topic of harvesting normalisation?

Q8. Your interest in seeing the Australian Research Online harvested data without normalisation?

ANDS Metadata

- Margaret (ANDS) provided an overview of the ANDS metadata, RIF-CS.

- RIF-CS (Registry Interchange Format Collections and Services) – Based on ISO2146.
http://ands.org.au/resource/techdocs.html#rifcs

- ISO 2146 (Registry Services for Libraries and Related Organisations) is an international standard currently under development by ISO TC46 SC4 WG7 to operate as a framework for building registry services for libraries and related organisations.

- Andrew (Monash) discussed his experience using RIF-CS in a Vital repository. The descriptions are at collection level not object level. Andrew is hoping to have some further information to share and show on this soon.

- Discussion on how the ANDS metadata caters to many groups wider than libraries and repositories.

Other metadata topics for future discussion

- Metadata standards for Creative Works.
Beyond the information provided at:


Topic Teleconferences and Including Creative Works in your Repository

Questions for Community

Q1. Any other repository managers thinking of using MODS in the future? Any information on your experience/research into moving to MODS? Community interested in hearing about it.

Q2. Do MACAR resource types need to be extended?

Q3. Could there be a loose agreement by Aus repository managers to adhere to DC types?

Q4. Could there be a loose agreement by Aus repository managers to make best effort (taking local metadata requirements in mind) to adhere to MACAR resource types?

Q5. The need to outline how we want the metadata to work, regardless of current software limitations – to help move beyond those limitations in the future. Your thoughts/comments?

Q6. Is there a need for a MACAR group be continued in 2009/2010 (with facilitation via CAIRSS)?

Q7. Your thoughts on the topic of harvesting normalisation?

Q8. Your interest in seeing the Australian Research Online harvested data without normalisation?

Links of possible interest

MACAR
http://macar.wikidot.com/

RUBRIC Toolkit Metadata Overview
http://rubric.edu.au/packages/RUBRIC_Toolkit/docs/Metadata_lite.htm
http://rubric.edu.au/packages/RUBRIC_Toolkit/docs/Metadata.htm

ANDS Metadata – Awareness Guide
http://ands.org.au/guides/metadata-awareness.html

ANDS RIF-CS
http://ands.org.au/resource/techdocs.html#rifcs

Australian Research Online
http://research.nla.gov.au/

MODS
http://www.loc.gov/standards/mods/.

Bibliographic Ontology Specification
http://bibliontology.com/

CAIRSS Teleconference notes on Creative Works
Teleconference Information
and
Including Creative Works in your Repository

CAIRSS SNAPSHOTS idea