Retrieving journals APC information

[Maxence Larrieu]

Dear Community,

From a DOI I would like to know if a publication has required APC, and if possible an estimation of the amount. How would you do ?

my ideas :

1. Use unpaywall API and if   oa_status is "hybrid" then APC is true

2. Use unpaywall API and if journal_is_in_doaj, query the DOAJ to get APC information (example )

3. Use openapc database

4. Some publishers share on the web a list of their journals with APC information (eg springernature) : if it does not exist already consolidate them and use this table.

What do you think ?

One problem I see is time ... APC information, and even journal business model [1], can change, whereas the 2nd and 4th solution are "snapshots", they are valid at the time they have been updated/created. A time based journal APC database would be the best.

Kind Regards,

Maxence

[1] Reverse flip open access journals : https://zenodo.org/record/2553582

[dan.p...@deltathink.com]

Hi Maxence, 

I have done a lot of work looking at APCs, so I hope the following helps.

I think your approach is probably the best using free/open sources. Note that OpenAPC tracks charges paid, which may include discounts, or specific deals for a paper. The DOAJ tracks "whatever list price the publisher states at the time" for the journal. (Apologies to either team if I have that wrong :-).

Consolidating publishers' price lists is a big job. They all use different formats. Some are better than others, but it's still a big data cleaning job (if you want more than one or two publishers). Often hybrid and fully OA journals are listed separately. Sometimes you need to ask the publisher for a structured list. Then you have the problem of matching DOIs to journals - e.g. where a print ISSN is used in a price list, but its electronic version is used in the index you're using. And that's assuming you have ISSNs to go on, which is not always true for all price lists. Also, some publishers have variable APCs, e.g. charge more for a CC BY license, so even with a comprehensive price list, you may not be able to predict the APC of a given paper. 

Part of my work has involved assembling a comprehensive list of APCs across the major publishers. We have to run the process annually to capture changes in APCs and journal status. We do this at the journal level so we can look at patterns across the market. It's a ton of work. I can't point you to an open data set…we're a small independent consultancy, and we don't have grant or foundation funding, so we have to charge access to cover our costs. Although we do make some stuff available openly [2nd chart], including analyses on pricing patterns. 

HTH,

Best,

Dan

[ric...@ourresearch.org]

Hi Maxence,

One gap to fill in is where oa_status is "gold", journal_is_oa is true, and journal_is_in_doaj is false. These are journals where we've seen that everything is bronze or hybrid and assigned gold status ourselves. These are probably best treated as hybrid when looking up APCs.

Richard

[Bianca Kramer]

Hi Maxence, Richard, all, 

The subset described by Richard is also the subset where it would be very interesting to find out which of these are diamond (no-APC) journals, that are not registered with DOAJ. 

For the study on diamond OA commissioned by cOAlitionS I am involved in, we are currently looking at ways to get a view on those non-DOAJ diamond journals. In a way, the flip side of Maxence's question :-) We attempted to collect information through crowdsourcing (https://tinyurl.com/diamond-journals) and also hope we can share many journal names that participated in our survey (500 of which were non-DOAJ diamond journals). But there still is a large unexplored area. 

kind regards, Bianca

Op vr 13 nov. 2020 om 18:32 schreef ric...@ourresearch.org <ric...@ourresearch.org>:

--
You received this message because you are subscribed to the Google Groups "Unpaywall discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unpaywall+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unpaywall/f43f578d-30ae-4965-9ceb-29ef7009dec0n%40googlegroups.com.

[Maxence Larrieu]

Hi every one,

thank you Dan, Richard and Bianca for your answer. It helped me a lot.

Because at least of transforming agreement, flipping journals and waivers, I understand that one cannot be sure that APC has been paid ( and a fortiori from a specific university).

However we can give some information about the evidence or not of APC. I have made an algorithm for that, here is the pseudocode : 

1. Is the DOI in OpenAPC ?

    Yes, return doi_in_openapc and the amount paid

    No, 2. is the journal inside OpenAPC and some APC has been paid the same year of the publication ?

        Yes, return journal_in_openapc and the average of the amounts paid over the year

        No, 3. is the publication in an hybrid journal ? (oa_status from unpaywall)

            Yes, return journal_is_hybrid

            No, 4. is the journal inside DOAJ list and has APC information ?

                Yes, return apc_journals_in_doaj, the price and the currency

                No, pass

the python code is here https://github.com/ml4rrieu/barometre_science_ouverte_uvsq/blob/main/b_enrichir_data.py#L159

for the 2nd step I have made a table openapc_journals.csv, please see https://github.com/ml4rrieu/open_science_tools/tree/main/apc_journal_tables

The people of DOAJ point me to this, which can be helpful (I have to include it for a 2nd version) ttps://github.com/lmatthia/publisher-oa-portfolios/

please feel free to comments,
Best Regards,
Maxence

[ric...@ourresearch.org]

Hi Bianca,

I'm sorry to say I lost track of this thread - but I wanted to let you know we'd like to help with the diamond OA study if we can. So as not to spam the thread, I'll send you an email with some of the raw data behind our Gold OA determinations and come back here with whatever seems useful.

Richard