Skip publisher if it equals JSTOR.

[lawlesst]

The publisher is being reported as JSTOR for some journals when JSTOR is probably the DOI minting agency. This just skips the publisher if it equals JSTOR. We could also consider not adding publishers for journal articles because the publisher information would be coming from another source.

[grahamtriggs]

Thanks Ted,

I'm 50 / 50 on whether we should worry about this, for two reasons.

1. We will be importing a large, curated journal set before "launch" and linking publications to that. In the case that the journal exists in the set, then we won't be creating journals or publishers anyway.

2. There is a question as to how much we should try to fix the data in the ecosystem. Mostly what I've done is normalize data (e.g. common representation of names to aid the guessing of authors). The one serious issue I dealt with was to ignore ISSNs that were errorneously included on books and chapters (because that made a total mess of the data).

Part of the outcome of this should be to highlight the problems in the ecosystem. Certainly, we need to track this - and the issues outlined above - and make it visible that these problems exist. Knowing that we can't automatically fix everything, we might want to accept that some of these things will be visible in the imported data anyway.

I'll speak to Mike later, and check that we are documenting the issues that we are working on - e.g. including it on the poster - or if we should leave the data to highlight the problem.

[lawlesst]

Both reasons make sense to me. Thanks for the explanation.