Print

---

Hi Francis,

Thanks for the response. Only after sending my email I remembered the SFM.
From their docs it looks like the SFM application only supports the 'normal' Twitter API with searches going back 7-9 days.
Applying for an application did change since the docs were published in 2015, as I have read in On Archivy [1]. Still, the approach of setting up SFM and allowing users to collect using their own credentials (presumably after a consultation with the library) looks good to me.

I will get in touch with DocNow folks when I do need to go further.

Regards,

Ben

[1]: https://medium.com/on-archivy/twitters-developer-policies-for-researchers-archivists-and-librarians-63e9ba0433b2


On 14-05-20, 13:30, "Code for Libraries on behalf of Francis Kayiwa" <[log in to unmask] on behalf of [log in to unmask]> wrote:

    Heya Ben,

    On 5/14/20 4:21 AM, Companjen, B.A. wrote:
    > Hello,
    > 
    > Following up on the thread about (not) getting tweets from the Twitter API older than 7 days, I would like to know if there is any experience providing access to the Twitter Premium APIs as an (academic) library. 
    > 
    > Various researchers have asked us at the Libraries how they could get tweets for their research projects and my answer has been pretty much the same each time: you would need to apply for API access (and accept the Terms of Use), understand that you can't get historical tweets, or pay.
    > I also point out that (depending on their research questions) Twitter may not provide a representative view of opinions and that they should consider whether it is ethical to use tweets. No one so far has replied with 'yes, all done and great, now can you help me further with the Premium API?' but I expect someone will at some point.
    > 
    > As the library manages licences for access to other datasources, I have wondered if we should and could provide managed access to the Premium APIs.
    > Although I have never applied for API access, I think individual researchers may have access to the API more quickly than when they need to apply to Twitter. But one person's misbehaviour could also impact all other researchers' projects. And I don't know what kind of costs would be involved.
    > 
    > I would like to hear if anyone has experience or thoughts about this. Feel free to respond off-list or on-list.


    That's interesting and reads a bit like GWU's Social Feed Manager[0]'s
    model. (or at least as I remember it). I am not sure if Laura et. al.,
    hangout here but you can for sure find them and other folks who may have
    Twitter Premium (sadly I do not) at the DocNow Slack[1]

    That said Premium or not if you elect to provide a service like this I
    would abstract this away from the library twitter account or whatever
    with an application like Social Feed Manager and/or the DocNow (full
    disclosure: I'm involved in the project) so you would be able to know
    who said bad actor is.

    Cheers,
    ./fxk


    [0] https://library.gwu.edu/scholarly-technology-group/social-feed-manager
    [1]
    https://docs.google.com/forms/d/1Wk0JdF2Cty2VHMqpf_QlJXVKQdUtfeeFhaYRben3qaM/viewform