Review of Aggregates Review Records #466
Replies: 81 comments 145 replies
-
[Please use this template for your feedback]:Filename: [For example] File Name: 7_CAM_AMTest_MReview.txt Pattern Logic: AND 6XX$x=Atlases [Case=Y] , OR 6XX$v=Atlases [Case=Y] Record LCCN (010): $a 74081983 Feedback: Are scientific atlases, and all things called atlases really aggregates? (spoiler: yes) |
Beta Was this translation helpful? Give feedback.
-
Good catch. I will have amended the pattern logic to eliminate descriptions of individual sessions, even though some of them could well be 'variety shows': Note that this results in zero hits in my small sample file, but other files might have records describing the entire radio program as a diachronic work. I did the same think for 9_CAM_AMTest_MReview.txt] This also results in zero hits in my small sample file. |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 31_CAM_AMTest_MReview.txt Pattern logic: AND 7XX Ind=X2=±t!!±5 [Case=Y] , NOWHERE 505$r, NOWHERE 505$a= / [Case=Y] LCCN of record: 73850836 Feedback: It's difficult to assess these based on 7XX Ind=X2's when so many of them appear twice in the output. Why is this happening with some records and not others? Is this a duplicate or bad data? From what I can tell all of these do look like aggregates. |
Beta Was this translation helpful? Give feedback.
-
Good catch. I have uploaded an updated version of the results file that shows the 7XX$e 22_AMTest_MReview (1).txt |
Beta Was this translation helpful? Give feedback.
-
Sorry about the repeated display of some fields, sometimes. I have provided an explanation in the general instructions section at the top of the page. |
Beta Was this translation helpful? Give feedback.
-
As we know, a 505 without a statements of responsibility (SOR) for different creators might be for chapters in a larger single work (i.e., a multipart work) and so is not a reliable indicator of an aggregating work. However, if a 505 is present (with or without SOR markers) **in conjunction with ** 240$k Selections, then the 'chapters' listed in the 505 are named parts of a multipart work and the record is describing a selection of those parts (i.e., an aggregating work). The presence of 7XXI2 = 2 acts in the same way as the 505, in this pattern match. Note that we can safely assume that a 505 in a record for a musical work (000/06=c||d||j) is describing different musical works in most cases (there are rare instances where they list the parts of a multi-part musical work). This assumption is why I 'except out' those records in this pattern match. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 33_CAM_AMTest_MReview.txt Record LCCN (010): 2003269135, 2008023086 Record LCCN (010): 2008023086, 2021050851 Record LCCN (010): 2009579581 Record LCCN (010): 2011535136 Record LCCN (010): 2012517286 Record LCCN (010): 2014018973 Record LCCN (010): 2014657512 Record LCCN (010): 2016417333 Record LCCN (010): 2017021183, 2020430404, 2021009375 Record LCCN (010): 2019001022 @tmqdeborah This rule seems solid. You might consider adding a count of "/: in the OR case of more than one to weed out some of these (if you only want collection aggregates). There are enough 2-expression aggregates to make it impractical to go beyond that. Some appear to be collection aggregates but cataloging seemed lacking: Record LCCN (010): 97046891 Record LCCN (010): 2002635132 |
Beta Was this translation helpful? Give feedback.
-
|
We can't help what weird things catalogers do. The question is whether the ISBD punctuation for statement of responsibility has been used correctly in the majority of the records found by the pattern; and it has. In larger files, this pattern finds thousands of records, the vast majority of which appear (from sampling) to be collection aggregates. I did try your suggestion of narrowing the search to 505$a that contain more than one ' / ', but that pulled too many records where one named work (not just an introduction or foreword) is provided in a list of what appear be chapters of a single work, i.e., two expressions of two works are embodied in the manifestation, i.e., the manifestation is a collection manifestation. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 34_CAM_AMTest_MReview.txt I'm not sure I have the experience with digital map files to say whether the 3 records found by this rule are for aggregates or not. The discs each have "DOQ digital orthophoto quadrangle data" for one specific geographic area (e.g. Douglas County, WI). Without viewing the contents of the disc I can't tell if it's a single navigable map, perhaps, or if multiple map views could be generated... Unless a cartographic data file is somehow an aggregate by definition (?), I'd like someone more familiar with this data format to weigh in on the usefulness of this matching rule. |
Beta Was this translation helpful? Give feedback.
-
|
I would think that more than likely that records that are coded as cartographic material and describe computer discs are likely to be describing manifestations that embody more than one expression (maps, etc.) I can't point you to anywhere that states this, but it is my understanding that databases are aggregating works. I have attached a larger file of examples (19): These examples include:
I, personally, think it is better to pull them to be considered later (Phase II) than include them in the Phase I non-aggregates file, but I will remove the pattern match so that we don't have to discuss it any further at this point. |
Beta Was this translation helpful? Give feedback.
-
|
@tmqdeborah - your call whether to remove the pattern. I just did a teeny bit of poking into what a "DOQ" is. https://pubs.usgs.gov/fs/2001/0057/report.pdf and it seems to be a single digital image derived from an overhead photograph of a geological area. I presume that the data on the CD-ROM would be accessed to produce this image with the right reader or software. That's as far as I have time to get into it. For me it just raises the possibllity that some (and potentially a large number if LC has records for all the DOQ's distributed by USGS on CD-ROM) of the items found by your algorithm could be considered non-aggregates.. Not sure I can suggest anything to add to the pattern that would avoid non-aggregates at this point - like I said, this needs map cataloger knowledge I don't have. |
Beta Was this translation helpful? Give feedback.
-
|
@tmqdeborah: is the assumption valid that 'computer disk' does not, in practice, cover floppy disks (of all sizes) and other disks with relatively low storage capacity? I imagine that Earth-sensing and telescope/spacecraft images in the early days were recorded on such disks, with one image per disk. |
Beta Was this translation helpful? Give feedback.
-
|
@GordonDunsire, the pattern is I remember the discussions about when to use disk vs. disc and I'm fairly certain that 'disk' was to be used for floppy disks and 'disc' was to be used for optical discs. So, my pattern is only looking for the larger discs that be for manifestations that embody more than one expression (more than one map, or what appears to be one map with data that changes what shows on the map/expressions of the map?) |
Beta Was this translation helpful? Give feedback.
-
|
35_CAM_AMTest_MReview.txt @lake44me See also: List245aCollectionTerms.txt |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 35_CAM_AMTest_MReview.txt I'm not comfortable assessing terms with none or only one or two results, even if all results are aggregates: Terms that seemed unreliable because they identified too many non-aggregates or questionable aggregates: articles: 1 (1 non-aggregate) Terms that identified aggregates or probable aggregates with no or few exceptions. cookbook: 30 (28 collection aggregates; 1 programming book that might be an aggregate; 1 directory of sources for cookbook collectors that might be an aggregate) |
Beta Was this translation helpful? Give feedback.
-
|
@lake44me, you have put a huge amount of work into this review. The bottom line is that this is somewhat of a scatter-gun approach to find plural words that indicate aggregating (collection) works. I agree that 'articles' is unfortunately used for 'legal articles' (e.g., 'articles of agreement') but there are many more instances where the term is for a collection of individual 'bibliographic articles'. I'm going to have to keep the questionable ones terms in the list; that way they are pulled for further analysis in Phase II instead of being lost in the non-aggregates file . |
Beta Was this translation helpful? Give feedback.
-
|
36_CAM_AMTest_MReview.txt @lake44me See also: List245aCollectionTerms.txt |
Beta Was this translation helpful? Give feedback.
-
|
Before you spend a lot of time on this, please see my response to #466 (comment) |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 36_CAM_AMTest_MReview.txt I haven't done the summary yet (may need to wait until tomorrow) but here is the file with annotations/terms and aggregate status counts: |
Beta Was this translation helpful? Give feedback.
-
|
@tmqdeborah Unfortunately I didn't see your reply, so analysis was just as detailed. Whatever you find useful, use... you might look for terms where there were no aggregates found, or the proportion was less than 50%. ,,, also using regex instead of word search led to false hits on several. I assume you are looking for any kind of aggregate including augmentation and parallel, but have noted where I thought these were possibilities in addition to collection ags. Lots of guesswork. |
Beta Was this translation helpful? Give feedback.
-
|
37_CAM_AMTest_MReview.txt @lake44me See also: List245cCollectionTerms.txt |
Beta Was this translation helpful? Give feedback.
-
|
Before you spend a lot of time on this, please see my response to #466 (comment) However, I have removed 'annotation' from the List245cCollectionTerms list (and the List245aCollectionTerms and List245bCollectionTerms lists), and I have added it to the List245cIntroductionTerms list. |
Beta Was this translation helpful? Give feedback.
-
|
Filename; 37_CAM_AMTest_MReview.txt I looked at unmatched terms in the list. I think these might be more likely to catch non-aggregates (books about something rather than containing something, or books that are single instance or uniitary work characterized by a type): a book of For other lists for rest of 245, "narratives" is a term I noticed a lot used in a modern sense in social science / athropology etc. works as a topic, not really aggregates of narratives. Not sure about the regexes for these - they didn't catch anything so may not be working or worth it. |
Beta Was this translation helpful? Give feedback.
-
|
38_CAM_AMTest_MReview.txt @lake44me See also: List520$aCollectionTerms.txt |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI will take this one |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 38_CAM_AMTest_MReview.txt Pattern Logic: AND 520$a=C:\Users\deborah.ASH\Documents\MarcReportTermLists\List245aCollectionTerms.txt [Case=Y] Feedback: I think that the 520 is an unsafe field to base aggregate pattern matching logic on. These are notes describing "aboutness", not "is-ness", and therefore none of these terms in the 520 $a indicate an aggregate. Further, I think the following fields should be removed from the list for consideration when it comes to other fields (such as 655) because they could represent singleton expressions when they occur throughout MARC: |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 39_CAM_AMTest_MReview.txt Other 2 seem more straightforwardly to be aggregates (proverbs, quotations). |
Beta Was this translation helpful? Give feedback.
-
|
The name of the file tells you the category of work that is the focus of the pattern matching. I should have provided this information earlier, but have just added it to the Introduction at the top of this page. So, the pattern that you quote is looking for a collection aggregate. "Arranged" is a problematic term (which is why I filtered out music records) and this is an example of it catching a record that is not for a collection aggregate. It is for an augmentation aggregate because of the supplementary content listed in the 245$c, and, hopefully, would be caught by a later 'AAM' pattern match. |
Beta Was this translation helpful? Give feedback.
-
|
40_CAM_AMTest_MReview.txt @lake44me The question for this review is whether it is most likely that a record with no 100 (creator person) but with one or more editors is describing a collection aggregate (contributions of expressions by various creators) |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 40_CAM_AMTest_MReview.txt I was suspicious of the: ones with "the editors of" in $c, though. For example, 010 $a 78089982 only 8 records with string "the editors of" in the whole 785 record batch so maybe not worth thinking about. I was going to suggested skipping those with 130 as well, but there was really only one (out of 4) that was questionable. 010 $a 75322844 when I looked at it (Internet Archive has it), yes it has the Revised Standard bible text but the bulk of it is commentary by A. L Moore (in the Introduction and in voluminous footnotes). Seems like this is an aggregate but a very interwoven one. |
Beta Was this translation helpful? Give feedback.
-
"Interwoven" in this case sounds like 'annotated' and annotations can either be minor (primary work augmented by minor work) or major (equal works: augmented + augmentation). This example (like "The annotated Alice in Wonderland") seems to me to be a collection aggregate in that the bulk of the content is the augmenting material. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI will take this one |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
@tmqdeborah Before someone starts on this - question that came up in a meeting - are all statistics considered to be aggregates? When are they not? If interspersed with text and interpretation...? If not about and RDA entity? |
Beta Was this translation helpful? Give feedback.
-
|
A collection of statistics can be analyzed by multiple people. One collection of statistics can be very different from another collection of similar statistics. A manifestation that includes a collection of statistics and an analysis of those statistics (to me) is a collection aggregate or (possibly) an augmentation aggregate depending on whether you consider the statistics or the analysis as 'primary' (a textual work with supplementary statistics). This and other categorization issues should definitely be discussed in Phase II. |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI will take this one |
Beta Was this translation helpful? Give feedback.
-
|
File name: 46_CAM_AMTest_MReview.txt Pattern logic: AND 6XX$v=Item In List (C:\Users\deborah.ASH\Documents\MarcReportTermLists\List6XX$v$xLCGFTterms+.txt) [Case=Y] Feedback: This goes back to the question again of what constitutes an aggregate. We should revisit the list this pattern logic uses when we revisit that discussion in more depth. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
@tmqdeborah This one is huge (197 records). Rather than posting individual comments, can I post the whole list with summaries for each term and results of my viewing of previews, tables of contents etc. with each individual record? |
Beta Was this translation helpful? Give feedback.
-
|
Filename: 47_CAM_AMTest_MReview.txt
Or looked for clues elsewhere in the record for other terms. There were 152 that I labeled "Aggregate" in the output of 197 records. I separately labeled Statistics, since the question about them is open (see 46_CAM_AMTEST..). There were 20 in the output. As we discussed with 48_CAM_AMTest review, form/genre terms that are plural but don't indicate an aggregate when they are applied to a single instance of the form should not be included. I found several indications of problematic terms like this: Not an aggregate, single instance of form/genre: 16 There were also some where matter in the form was included interspersed with other text, etc. and not discrete: 4 Some did not appear to be an aggregate, but to have the form as a subject: 3 -- Bibliography 2 Two I was not sure about.: LCCN: 2001131437 has $x Genealogy in subjects. But the title is From a little white church : $b the history of Fairfax United Methodist Church (formerly Duncan Chapel), Fairfax, Virginia . LCCN: 97034014 has $x atlases in subjects. There is a medical meaning of the term "atlas" that is not a collection of maps. It might have something to do with a comprehensive reference resource used for determining diagnoses of diseases based on factors, which is what this is. I also know of another work, "Linguistic atlas of the Gulf States" which was a project at Emory for systematically recording and analyzing word and pronunciation variations for different parts of this geographic area, based on recordings of local residents speaking. Both of these might fit an "aggregate" definition but not in the way expected. All of this is guesswork, but gives a picture. @tmqdeborah If you want more specific examples I can add them. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Adding some detail ... The first examples discussed in the meeting were bibliographies (and catalogues raisonnes). These are metadata description sets because they are structured descriptions of RDA entities (e.g. works). The relevant Toolkit Glossary definitions are: metadata description set: metadata statement: A metadata work that is a piece of metadata that assigns a value to an RDA element that describes an individual instance of an RDA entity. metadata work: A work that is a metadata statement or a metadata description set. Dictionaries may, or may not, metadata description sets. E.g. A biographical dictionary is, but a word dictionary composed by a single agent (e.g. Samuel Johnson) is not. The second issue we discussed is the application of plural forms to single instances. This is a consequence of subject heading conventions that favour the plural to describe the class, e.g. 'lions' vs 'lion'. Thus 'menus' will be applied to a single mehu manifested on a card and a set of menus manifested by a volume or online resource. The former is not an aggregate; the latter is a collection aggregate. Some entries in the list(s) provided by @lake44me are aggregates irrespective of the number of WEMs to which they are applied (e.g. metadata description sets). These are 'safe' indicators of an aggregate WEM. However, I think if we look beyond printed volumes it is easy to find examples of single instances (non-aggregates) of WEMs:
I therefore think it is not safe to check terms in field 655 to determine 'aggregateness' and the assumption should be that all terms in 655 cannot distinguish an aggregate from a non-aggregate because the plural form is always used. In my opinion, field 655 is a dead-end as a test and there is no need for further work on it. |
Beta Was this translation helpful? Give feedback.
-
|
I think it is safe to use both 655 and 650 $v for things that are aggregates whether they are single instances of the class or not, but not for anything else. I thought we were in agreement on that point. If the work only extends to these types of aggregates, would you agree that it was a worthwhile endeavor @GordonDunsire ? For genres/forms/types of things that are not aggregates when they occur as single instances, I agree with Gordon that we cannot use 655 to find them. We may be able to use 650 for some of them, though, such as $v Translations in LCSH, which is ONLY supposed to be used for collections of translated works. Most free-floating subdivisions that can be used for collections can also be used for individual works, though, I think, and those wouldn't be candidates for inclusion in our list. @AdamSchiff is the expert on that but it's definitely figure-out-able in my opinion and I don't know if I would say it was a dead-end as a test. |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI , I'm now coming around to Gordon's view - there probably aren't genre terms where all instances are aggregates by definition, because they may or may not be "metadata works" referring to potential RDA entities (although Bibliographies sounds close, but bibliograhies take different forms and organization. some that don't fit with the idea of a "metadata work" well). The use of rules to examine treatment of various genres of works or expressions might be good for post-conversion review by users of our transformation, but I don't think we should use them to make the cut between "Phase I" and "Phase 2" transformation. I can plow ahead with the large files to identify the percentage of cases when a single non-aggregated instance of a genre/form is being described, but based on what I've seen so far, there are too many of these among non-"metadata work" categories. |
Beta Was this translation helpful? Give feedback.
-
|
How is a collection of recipes or bibliographies or puzzles, etc., 'by' a single creator different from a collection of poems, short stories, novels, etc. (other than that they might be shorter)? I removed a few of the more questionable terms from the ListV650aLCGFTterms+.txt list and have put all of the entire set of Item lists in a new folder Aggregate markers Term lists_20241106 I added the filters 'NOT $x NOT $vFiction' to eliminate when a 650 might be for a single work, rather than a collection and, going by the examples, it seems to work quite well. Or at least well enough that it seems (to me) to be better to pull the records so that they can be further refined (if possible) in Phase II than to let them go through and be lost in the Phase I file. |
Beta Was this translation helpful? Give feedback.
-
|
I think that's sufficient and agree with @tmqdeborah , for what it's worth. And think we can refine post-transform. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 49_CAM_AMTest_MReview.txt These were the only 4 records identified as aggregates and all of them appear to be at least worthy of consideration to treat as collection aggregates. I'm going into detail because I was about to write this logic rule off as not useful until something told me to try to look outside of the MARC record for more info. Guess that's the only way to know for sure, darn it. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks for writing this up @lake44me. I appreciate hearing that your research concurs with mine. The logic behind this pattern is that, in my experience (from doing outside (Amazon, Internet Archive) research), most records:
The same kind of research/thinking applies to a number of other patterns in the list. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 50_CAM_AMTest_MReview.txt Record LCCN (010): $a 99178996 Feedback: Want to say this is an aggregate but at 197 pages really not sure Plus generally there are some 130s in this file |
Beta Was this translation helpful? Give feedback.
-
|
I should have added that I have applied the NLNZ Aggregate Guidance to my files. "What kind of aggregate manifestation do you have? Most of the resources collected by the NLNZ are aggregate manifestations. This includes any resource which contains illustrations or other supplementary content. To help decide which kind of aggregate manifestation you have, you need to decide how many works are embodied in the manifestation, and then which are main works, and which are supplementary." |
Beta Was this translation helpful? Give feedback.
-
|
The theory behind this pattern match is that if there is no 100, 110, 111, and 245$c begins with 'edited' or 'editor' the likelihood is great that the editor is an editor of compilation. Most of my sampling of the results seems to bear this out, even for most of the ones with 130's. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 129_DW_LC50thTest_MReview.txt Record LCCN (010): $a 94061899 Record LCCN (010): $a 94140168 Record LCCN (010): $a 95070204 Record LCCN (010): $a 98032162 Record LCCN (010): $a 2016202749 |
Beta Was this translation helpful? Give feedback.
-
|
All loose-leaf and I can't recall any non-diachronic. As this is a tad shorter file I'll do a quick relook once I've finished the other files |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
Filename: 135_PAM_LC50thTest_MReview.txt Feedback: this pattern logic picks up records with two or more language expressions for monographs and serials. Is it intended to include serial records? |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
File name: 138_PAM_LC50thTest_MReview.txt Pattern Logic: NOT 546$a='^The [a-z][a-z]+ work||^In [A-Z][a-z]+ only||^In [A-Z][a-z]+ with||^In [A-Z][a-z]+, with||^In [A-Z][a-z]+;||in [A-Z][a-z]+ only||in [A-Z][a-z]+, with||in [A-Z][a-z]+;||in [A-Z][a-z]+ with' [Case=Y] [RegEx] , AND (SAME TAG) 546$a='^Text in||^Texts in||^Songs in||^Sung in' [Case=Y] [RegEx] , AND 041=(ąa.*ąa|ąd.*ąd) [Case=Y] [RegEx] , OR 041=(ąa[a-z][a-z][a-z][a-z]|ąd[a-z][a-z][a-z][a-z]) [Case=Y] [RegEx] LCCN: 2012593511 Feedback: These language patterns relying on the contents of 041 and 546 rely on MARC records representing multiple expressions rather than a single expression expressed using multiple languages (a famous example, War and Peace, in Russian and French in its original expression). Most of the time, I think the patterns are fine because 99.99% of the time, they find multiple expressions. Maps, though...I think single maps (not atlases) are maybe special and should be excepted from all of these logics. Text in multiple languages on a map is pretty common, and I don't think it makes multiple expressions. I think it makes them like War and Peace. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
The LCSH free-floating subdivision $v Atlases is not used on cartographic works, it's only used on scientific works.
The scope note: Use as a form subdivision under scientific and technical topics for works consisting of comprehensive, often systematically arranged collections of illustrative plates, charts, etc., usually with explanatory captions.
I would not assume that the presence of $v Atlases or $x Atlases (if the previous coding was never changed after $v was invented) would generally mean an aggregate unless each illustration were considered a separate work or the captions were considered separate works from the pictures.
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Laura Akerman ***@***.***>
Sent: Tuesday, November 12, 2024 8:50 AM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] Review of Aggregates Review Records (Discussion #466)
Filename: 47_CAM_AMTest_MReview.txt
Pattern logic: AND 6XX$x=Item In List (C:\Users\deborah.ASH\Documents\MarcReportTermLists\List6XX$v$xLCGFTterms+.txt) [Case=Y]
Feedback: I quickly labeled these , usually making the assumption that certain terms were aggregates by definition:
* Congresses: (if denoting a conference publication)
* Conversation and phrase books
*Catalogs
*Dictionaries
*Directories
*Encyclopedias
*Exhibitions (denoting an exhibition catalog)
*Gazetteers
*Glossaries vocabularies etc:
*Guidebooks
*Indexes
*Literary collections
*Patterns
Or looked for clues elsewhere in the record for other terms.
There were 152 that I labeled "Aggregate" in the output of 197 records.
I separately labeled Statistics, since the question about them is open (see 46_CAM_AMTEST..). There were 20 in the output.
As we discussed with 48_CAM_AMTest review, form/genre terms that are plural but don't indicate an aggregate when they are applied to a single instance of the form should not be included. I found several indications of problematic terms like this:
Not an aggregate, single instance of form/genre: 16
-- Case studies 5
-- Personal narratives 3
-- Sermons 2
-- Commentaries 6
There were also some where matter in the form was included interspersed with other text, etc. and not discrete: 4
-- Case studies 3
-- Anecdotes 1
Some did not appear to be an aggregate, but to have the form as a subject: 3
-- Bibliography 2
-- Case studies 1
Two I was not sure about.:
LCCN: 2001131437 has $x Genealogy in subjects. But the title is From a little white church : $b the history of Fairfax United Methodist Church (formerly Duncan Chapel), Fairfax, Virginia .
I didn't find this to examine but I assume the church history might include some genealogical background on individuals but not be entirely composed of it.
I observed several other records where $x Genealogy was applied, but the content was data like a census index, hand-copied records of various kinds, etc. that could likely be useful in genealogy research but weren't genealogies, so I suspect this was a practice at The Library Of Congress at some time in the past.
LCCN: 97034014 has $x atlases in subjects. There is a medical meaning of the term "atlas" that is not a collection of maps. It might have something to do with a comprehensive reference resource used for determining diagnoses of diseases based on factors, which is what this is. I also know of another work, "Linguistic atlas of the Gulf States" which was a project at Emory for systematically recording and analyzing word and pronunciation variations for different parts of this geographic area, based on recordings of local residents speaking. Both of these might fit an "aggregate" definition but not in the way expected.
All of this is guesswork, but gives a picture. @tmqdeborah<https://urldefense.com/v3/__https://github.com/tmqdeborah__;!!K-Hz7m0Vt54!ibySNsdfvCqi-7BNfkFr7Xu_6zeyPT44HFrKT6UtEG87qHII0VWTI5PuDv7JFKcHktZZCddrT4i0d3TDgwfZ8h0$> If you want more specific examples I can add them.
—
Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/466*discussioncomment-11229310__;Iw!!K-Hz7m0Vt54!ibySNsdfvCqi-7BNfkFr7Xu_6zeyPT44HFrKT6UtEG87qHII0VWTI5PuDv7JFKcHktZZCddrT4i0d3TD9jAC4-Y$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVB6VKOMIG3RJJJQKY632AIWUVAVCNFSM6AAAAABPZZZ4UGVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCMRSHEZTCMA__;!!K-Hz7m0Vt54!ibySNsdfvCqi-7BNfkFr7Xu_6zeyPT44HFrKT6UtEG87qHII0VWTI5PuDv7JFKcHktZZCddrT4i0d3TDNh22FcQ$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
"collections of illustrative plates, charts, etc., usually with explanatory captions" ...this relates to the first thing I was asking Deborah about w/r/t "AND 6XX$x=Atlases [Case=Y] , OR 6XX$v=Atlases [Case=Y]" logic. "Are all atlases (non-geographic included) aggregates?" @tmqdeborah See Adam's comment here about this. It was my first thought as well, the collections of photographs/plates/etc. in a scientific atlas, for instance, aren't really rising to the level of encyclopedic entries etc. because I think they're usually by the same person on a common theme(ish), right @AdamSchiff ? |
Beta Was this translation helpful? Give feedback.
-
|
Unless there are chapters by different persons, I wouldn't expect a scientific atlas to be an aggregate, unless every illustration is considered a separate work, which I suppose it could be. For example in an anatomical atlas, you'd have microscopic photographs of slices of various tissues of the body, perhaps showing the difference between healthy and diseased states. Are these photos really each separate works? Are the captions for each photo a separate work from the photo?
Adam
Adam L. Schiff
Principal Cataloger
University of Washington Libraries
(206) 543-8409
***@***.***
…________________________________
From: Crystal Yragui ***@***.***>
Sent: Tuesday, November 12, 2024 5:43 PM
To: uwlib-cams/MARC2RDA ***@***.***>
Cc: Adam L Schiff ***@***.***>; Mention ***@***.***>
Subject: Re: [uwlib-cams/MARC2RDA] Review of Aggregates Review Records (Discussion #466)
"collections of illustrative plates, charts, etc., usually with explanatory captions" ...this relates to the first thing I was asking Deborah about w/r/t "AND 6XX$x=Atlases [Case=Y] , OR 6XX$v=Atlases [Case=Y]" logic. "Are all atlases (non-geographic included) aggregates?" @tmqdeborah<https://urldefense.com/v3/__https://github.com/tmqdeborah__;!!K-Hz7m0Vt54!n8U4AZxVcm2HnCh43J_Ig9uwJ7Mz6NiQ_2liIu5V46qPy4Au4NLyKA20EwZ_rbybmqmuCL2hR5Q0eb1aQiL_gBE$> See Adam's comment here about this. It was my first thought as well, the collections of photographs/plates/etc. in a scientific atlas, for instance, aren't really rising to the level of encyclopedic entries etc. because I think they're usually by the same person on a common theme(ish), right @AdamSchiff<https://urldefense.com/v3/__https://github.com/AdamSchiff__;!!K-Hz7m0Vt54!n8U4AZxVcm2HnCh43J_Ig9uwJ7Mz6NiQ_2liIu5V46qPy4Au4NLyKA20EwZ_rbybmqmuCL2hR5Q0eb1ao3dUcNA$> ?
—
Reply to this email directly, view it on GitHub<https://urldefense.com/v3/__https://github.com/uwlib-cams/MARC2RDA/discussions/466*discussioncomment-11234107__;Iw!!K-Hz7m0Vt54!n8U4AZxVcm2HnCh43J_Ig9uwJ7Mz6NiQ_2liIu5V46qPy4Au4NLyKA20EwZ_rbybmqmuCL2hR5Q0eb1adAM-4ws$>, or unsubscribe<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/ADFBVBZJHK3DKD2VXAO6ROL2AKVFBAVCNFSM6AAAAABPZZZ4UGVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTCMRTGQYTANY__;!!K-Hz7m0Vt54!n8U4AZxVcm2HnCh43J_Ig9uwJ7Mz6NiQ_2liIu5V46qPy4Au4NLyKA20EwZ_rbybmqmuCL2hR5Q0eb1aIMtats4$>.
You are receiving this because you were mentioned.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
I don't know if we want to consider them that way for the purposes of this project. I don't think they are separate works either. Does the RSC? |
Beta Was this translation helpful? Give feedback.
-
|
@CECSpecialistI and others: From a conceptual pov, if the content embodied by a manifestation is analysed and separated into identifiable, stand-alone, and clearly bounded sections, then each section is an expression. A piece of a jigsaw puzzle is not an expression because it fails the identifiability and stand-alone tests. A paragraph of text in the middle of other paragraphs is not an expression, even if it is numbered or captioned, because it fails the stand-alone test. The image of a discrete object embedded in a painting is not an expression because it fails the clearly-bounded test (there is no separation by nothing/space from the rest of the image). If an expression is determined, it must be a realization of one and only one work. From a practical pov, an application of RDA must choose which expressions or works to describe, and the method of description. The options are the same for all entities related to a manifestation:
The immediate answer is that the RSC would apply the conceptual pov only, and would consider every still image to be a distinct work/expression, including maps in atlases, photographs in photography books, etc. This is irrespective of who creates them. The question of single authorship is only a factor for determining identifiability and stand-alone criteria for certain kinds of content. As previously mentioned, Johnson's dictionary is not treated as an aggregate because the entries do not stand alone and are known to be by a single agent. This reasoning would apply to other dictionaries or encyclopedias where the agent is a collective agent with a singular pov (beyond the content). From a practical pov, the main purpose of identifying the presence of one, or more than one, expression in a manifestation is to determine how to interpret the main entry fields and identify the work (or expression) as an aggregating work or not. This also determines the transform of added entry fields. The method of making the identification is a mix of the conceptual analysis and the practical analysis; the latter because we are dealing with existing, legacy data. The main impact of identifying an aggregate manifestation is the balance between minting an opaque IRI for the work and not attempting to de-duplicate, or minting a de-duplication IRI based on the main entry values. For example, an augmentation aggregate cannot be de-duplicated from the main entry data alone, because the main entry will be for the predominate work, not the aggregating work. A secondary impact of identifying an aggregate manifestation is that the main entry fields are assumed to refer to the aggregating work. In that case, the transform should not mint an (aggregating) expression based on the work. This is easily avoidable by not minting an expression based on the main entry data (or minted work). Are we getting too deep in the weeds? |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
Here is a place where we can provide feedback for Deborah's Aggregates Review files.
=======================================================
Please follow the template below to start a new comment for each review question that you have, but please use REPLY to reply to questions.
=======================================================
The files for review are in the New 3_AMTest MReview results folder.
Please open an MReview.txt file, e.g., 7_CAM_AMTest_MReview.txt, and:
look at the pattern logic provided in the heading, e.g., AND 6XX$x=Atlases [Case=Y] , OR 6XX$v=Atlases [Case=Y]
Note that the name of the file tells you the category of work that is the focus of the pattern matching:
look at a sample of the records that matched that logic, e.g.:
010 $a 74081983
110 1 $a SKF Industries, inc. $b Technology Center.
245 10 $a Rolling bearing damage : $b a morphological atlas / $c [compiled from the photographic archives of the SKF Industries, inc. Technology Center] by T. E. Tallian ... [et al.].
300 $a 351 p. : $b ill. ;
504 $a Bibliography: p. 349-351.
650 0 $a Roller bearings.
650 0 $a Ball-bearings.
650 0 $a Metallography $x Atlases.
650 0 $a Metallography $x Atlases.
700 1 $a Tallian, T. E.
provide feedback about whether the percentage of false hits vs. good hits in a file means that we cannot use a pattern
if you can think of a way to tweak a pattern to get better results, please provide it
if you can think of any patterns that are not provided in the spreadsheet, please suggest them
bear in mind that some of the discussions about what constitutes an aggregate might have to wait until Phase 2 of the project
note that some displays of records show fields appearing twice; this is a limitation of our program because it cannot tell whether a field that I have chosen to display as a default is also displaying because it was used in a pattern match.
SEE ALSO the explanatory document at Aggregate Markers Documentation.20241014.docx
Beta Was this translation helpful? Give feedback.
All reactions