Skip to content
This repository has been archived by the owner on Jun 23, 2020. It is now read-only.

dct:title of bibo:Journal not generated correctly for RDA records #762

Closed
acka47 opened this issue Dec 9, 2015 · 9 comments
Closed

dct:title of bibo:Journal not generated correctly for RDA records #762

acka47 opened this issue Dec 9, 2015 · 9 comments
Assignees

Comments

@acka47
Copy link
Contributor

acka47 commented Dec 9, 2015

[Edit: Reported by edoweb customers on 2015-12-02.]

See e.g. http://lobid.org/resource/HT016320444 (pre-RDA record) vs. http://lobid.org/resource/HT018819975 (RDA record).

Frrom the source of HT016320444:

<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">Q</subfield>
</datafield>
<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">I</subfield>
  <subfield code="a">Öffentliche Klärschlammentsorgung ...</subfield>
  <subfield code="n">Elektronische Ressource</subfield>
</datafield>
<datafield tag="310" ind1="-" ind2="1">
  <subfield code="a">Statistische Berichte Rheinland-Pfalz / Q / 1</subfield>
</datafield>
<datafield tag="331" ind1="-" ind2="1">
  <subfield code="a">Statistische Berichte Rheinland-Pfalz</subfield>
</datafield>

As you can see, the field 310 contains information that is already present in 331 and fields 085. I guess, witht he switch to RDA they decided to get rid of this redundancy (which is a good thing, I think).

In the relevant parts of the RDA record's source (HT018819975) 310 is missing:

<controlfield tag="030">a||dr||||||27</controlfield>
<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">O</subfield>
</datafield>
<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">II</subfield>
</datafield>
<datafield tag="331" ind1="-" ind2="1">
  <subfield code="a">Statistische Berichte</subfield>
</datafield>

Somehow we will have to generate a correct dct:title both for pre-RDA and for RDA journal records

@acka47 acka47 added the bug label Dec 9, 2015
@acka47 acka47 self-assigned this Dec 9, 2015
@acka47
Copy link
Contributor Author

acka47 commented Dec 9, 2015

The relevant part of the current morph currently only take into account 310, 331 and 333 for generating dct:title and work something like this:

  • They apply to all types of resources
  • dct:title is generated from 310 if it existsts. 331 and 333 are used as otherTitleInformation.
  • If 310 doesn't exist, 331 is used for dct:title and if 331 doesn't exist, we use 333.

I see different possibilities for solving this issue:

  1. We could do a Journal-specific transformationto for dct:title by concatenating 085s to 331.
  2. We could adjust the current global transformation by
    a) adding a rule "concatenate 085 when 331 is used for generating the title"
    b) adding a rule "if 085 exists, create dct:title by concatenating 331 and 085".

Probably 2.b) would be the easiest solution but it would mean that the title for bibo:Journals are generated differently pre- and post-RDA.

What do you think, @dr0i?

@acka47 acka47 assigned dr0i and unassigned acka47 Dec 9, 2015
@acka47 acka47 added working and removed working labels Dec 9, 2015
@acka47
Copy link
Contributor Author

acka47 commented Dec 9, 2015

Our colleague Stephani pointed to the fact that the information to be concatenated to 331 in RDA records should be taken from 360 (which seems to be redundant to 85) :-/.

@dr0i dr0i added the working label Dec 10, 2015
@dr0i
Copy link
Member

dr0i commented Dec 10, 2015

As both your examples have also 360 and this field is even mentioned in the mb-dnb definition (whereas 085 is not mentioned) I would suggest to substitute 085 with 360 and execute your propoal "2.a", resulting in

We could adjust the current global transformation by a) adding a rule "concatenate 360 when 331 is used for generating the title"

Or do we need the more detailed subfields differentiating 085 from 360? Or, not sure about that, is it even possible that 085 exists while 360 does not?

@acka47
Copy link
Contributor Author

acka47 commented Dec 10, 2015

Having looked at the data again, using 360 is not feasible as it results in title fields that aren't requested by stakeholders. See for example HT018785916 (source):

<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">K</subfield>
</datafield>
<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">VIII</subfield>
</datafield>
<datafield tag="085" ind1="x" ind2="1">
  <subfield code="l">1</subfield>
  <subfield code="a">Einrichtungen für ältere Menschen in Bayern ... und ambulant betreute Wohngemeinschaften in Bayern ...</subfield>
</datafield>
<datafield tag="331" ind1="-" ind2="1">
  <subfield code="a">Statistische Berichte</subfield>
</datafield>
<datafield tag="360" ind1="-" ind2="1">
  <subfield code="a">K</subfield>
</datafield>
<datafield tag="360" ind1="-" ind2="1">
  <subfield code="a">VIII</subfield>
</datafield>
<datafield tag="360" ind1="-" ind2="1">
  <subfield code="a">1, Einrichtungen für ältere Menschen in Bayern ... und ambulant betreute Wohngemeinschaften in Bayern ...</subfield>
</datafield>

Generating the title from 331+085.l the structure provided by subfields enables to provide what stakeholders expect: Statistische Berichte / K / VIII / 1
Using 331+360 we are missing structure an can only provide Statistische Berichte / K / VIII / 1, Einrichtungen für ältere Menschen in Bayern ... und ambulant betreute Wohngemeinschaften in Bayern ... – which stakeholders would not be happy with.

Thus, we will do it as I said in #762 (comment), 2. a) using only subfield l of 085: "concatenate 085.l with "space-slash-space" (/) when 331 is used for generating the title"

dr0i added a commit that referenced this issue Dec 11, 2015
See #762.

- fix also a redundant title entry, see BT000003404
- add HT018785916 to the test set
- update tests
dr0i added a commit that referenced this issue Dec 11, 2015
See #762.

- fix also a redundant title entry, see BT000003404
- add HT018785916 to the test set
- update tests
dr0i added a commit that referenced this issue Dec 11, 2015
See #762

- fix also a redundant title entry, see BT000003404, see #761
- add HT018785916 to the test set
- update tests
@dr0i
Copy link
Member

dr0i commented Dec 14, 2015

Deployed to staging.

@dr0i dr0i assigned acka47 and unassigned dr0i Dec 14, 2015
@dr0i
Copy link
Member

dr0i commented Dec 14, 2015

Removed the bug label because it's no bug. Everything worked as specified. It's rather a feature request, an enhancement.

@acka47
Copy link
Contributor Author

acka47 commented Dec 14, 2015

+1

@acka47 acka47 added deploy and removed review labels Dec 14, 2015
@dr0i
Copy link
Member

dr0i commented Dec 14, 2015

Deployed to production, closing.

@dr0i dr0i closed this as completed Dec 14, 2015
@dr0i dr0i removed the deploy label Dec 14, 2015
dr0i added a commit that referenced this issue Jan 11, 2016
See #762

- fix also a redundant title entry, see BT000003404, see #761
- add HT018785916 to the test set
- update tests
@acka47
Copy link
Contributor Author

acka47 commented Feb 16, 2016

This is a sub-issue of hbz/lobid#161.

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants