Journal abbreviations

One of the things the folks from the Smithsonian that I talked to wanted was a web service that would resolve journal abbreviations.  Actually, what I think they really want is a service that can take in a cryptic citation and output a nicely formatted one with the full journal title, etc.  There are lots of web sites that list various journal abbreviations, and even All That JAS which is a bibliography of those sites, but in my limited looking around I didn't find anything that quite does what they want, even for just the journal abbreviations.

I imagine that many OpenURL resolvers try to cope with journal abbreviations, and a citation-expander would fit easily into the OpenURL 1.0 framework, but I'm not aware of one.  Not having a handy list of journal abbreviations, I thought it would be interesting to see what could be gleaned from WorldCat.

We found nearly 300,000 bibliographic records with a 210 field (abbreviated title), most of which also had a 222 field (key title).  From these we found more than 12,000 abbreviations which correspond to 46,000 different words.  For example, the three letter string mil was used as an abbreviation (after NACO normalization) for:

• military
• militar
• militaire
• militarrundschau
• militare
• milscap
• militmilitariss

After duplicate journal names were eliminated we found about 200,000 journal abbreviations with corresponding full journal names.  Records like #09533880 show that Kōgyō kayaku is equivalent to:

• Journal of the Industrial Explosives Society, Japan
• Kōgyō Kayaku Kyōkai shi
• Explosion and Explosives
• Expolosion, explosives and pyrotechnics

and, if I'm reading the record correctly, also to:

• 工業火藥
• 工業火藥協會誌

There's lots of information about journal names in WorldCat.  We've discussed the idea of doing an xISSN service similar to xISBN.  Is there is a place for partial citation decoding in that?  Or has someone already done this?

Thanks to Tom Dehn, who did most of the data crunching.

--Th

Comments

[Todd Holbrook]

I'm the author of CUFTS - an open source fulltext link resolver and electronic journal management system. A web service (or downloadable database) of journal abbreviations would be very useful, as would something like xISSN. We rely on title lists from the vendors to build the knowledgebase, and often run into issues with bad or incomplete data that these services could help reduce.

For example, we often run into the case where one vendor will provide the eISSN and another the print for the same journal. If we had a way to easily link those, we could make sure the system matched on both regardless of which one the user linked in with.

[Bryan Campbell]

It be worthwhile to talk to someone at Thomson Gale, which publishes the 2 volume annual Periodical Title Abbreviations (v.1 By abbreviation, v.2 By title).

[Keith Jenkins]

From working at the reference desk, I know that the problem of journal abbreviations is one that researchers are constantly struggling with. However, this problem is not limited to journals, but also includes conference proceedings, government reports, etc. There are some great resources out there that attempt to connect abbreviations and full titles, but sometimes a user finds a citation that uses its own unique abbreviation and will never be found in a standard list.

Working from the idea that we can never complete a list of every abbreviation that will ever be used for a given title, I decided to try a different approach: starting with the abbreviation in hand, what full titles might this abbreviation possibly represent? I built a prototype, called JAbbr, that uses a list of titles from the Cornell library catalog to offer a list of possibilities. Users can then click on a title in the list to view the catalog record. JAbbr can be found at:

http://alteriseculo.com/jabbr/

While JAbbr has proven to be incredibly useful, there still are significant limitations to its approach. Due to the lenient matching algorithm, there is often more than one possible match, which requires the user's judgement to select the most appropriate result. However, this human intervention will always be necessary in some cases, since many shorter abbreviations are re-used by different communities. For (an extreme) example, CQ may refer to any of "California quarterly", "Carnegie quarterly", "Congressional quarterly", etc. Choosing from these options would require knowing the context of the citation, and may also require a fair amount of either trial-and-error or luck.

Even though JAbbr is based upon a list of over 300,000 titles, another shortcoming of the current prototype is that it only includes serial titles held at a single institution. The next logical step would be to incorporate even more titles into JAbbr. (Perhaps from OCLC's WorldCat?)

In the end, I think a combined approach to deciphering journal abbreviations would work best: a list of unique and known abbreviations could be used to automatically handle 98% of requests, while ambiguous or novel abbreviations would require some sort of human intervention before completing the request.

-Keith

[Thom]

JAbbr is great! I think I've seen it before, but forgot about it. Of course, within an OpenURL context, you probably would have some year and volume information that could often resolve the ambiguities. If you had a good registry of actual articles it would be even easier.

--Th

[Peter McCracken]

I manage data for Serials Solutions. We realized early on that we couldn't rely on ISSN, since different versions of the same journal have different (or no) ISSN. We created a work-level identifier when we created our database, and create a new identifier whenever we add a new resource. Since we have an identifier that isn't the ISSN, we can accept and track print ISSN, eISSN, previous ISSN, mistyped ISSN, etc., for each resource. A source can send a bad ISSN (because that's all it's got) and we'll know what they meant to say. Without that ability, a link resolver won't be able to figure out if the target exists or not.

In addition, we take in many different versions of titles from the CONSER MARC records we use, including the 210 field (abbreviated titles). With our very-recently-released 'alternate title searching' functionality, you can type in an abbreviation and get the actual title(s) associated with that abbreviation. So, for example, type in "j acad" and you might get back "Journal of academic librarianship", "Journal of the Academy of Management," and "Journal of the Academy of Marketing Science." (Of course, results will depend on the titles your library can access.)

That said, I don't understand why we still use journal abbreviations. They were great when we only had the limited space on a card in a card catalog, or a publisher had to to limit the number of pages in Chem Abstracts, or something similar - but today, it's just a bunch of 1s and 0s. Who but the initiated will know what "J Acad Mark Sci" stands for? Who's going to know to type "J acad librariansh" instead of "J acad lib" or "J acad libnshp"? Cutting off the last two letters hardly seems to qualify as an 'abbreviation' to me. Why do we continue to intentionally confuse our patrons?

[john]

I think a combined approach to abbreviations would work best: a abbreviations could be used to automatically handle requests, while novel abbreviations would require human intervention before completing the request.