Thread 2: A corpus-assisted discourse analysis of linguistic transphobia on Mumsnet

[VitoCorleoneOfMNMafia]

At fewer than 20 posts left, time for part two.

Thread 1: https://www.mumsnet.com/talk/womens_rights/5057460-a-corpus-assisted-discourse-analysis-of-linguistic-transphobia-on-mumsnet

Site Stuff thread that tipped off MNHQ: https://www.mumsnet.com/talk/site_stuff/5057903-mumsnet-corpus

Oh go on then

Another one for the AI generator - a bride in full veil, frock and bouquet combo and a six foot Tunnock's wafer in a tailcoat.

Placemarking

Me, me! I said that! Fame at last 😊

I was having a serious thought that led to a whimsical one.

Serious- could we get AI summaries of threads so people unable to keep up can stay in the loop? Court cases and threads like this would be massively more accessible if we could.

Whimsical- what the hell would AI make of tunnocks , cheese, BoiledBeetleverse and random references to our local copper, monitors etc?

<HappySigh>

The occasional vigorous objections to threads wandering sideways and poster cheering each other with cakes, jokes, and cheese have always made me wonder whether the objectors had even met a human.

If the objectors were in the business of training non-human tech that can't cope with human methods of communication it all makes sense.

within the MN ‘sandbox’ ie were they trying to \ did they establish who was the real life person posting?

No I don't think so, I think their goal was to test their authorship methods using the data of posters who had posted a number of times.

I am wondering why the decision to anonymise the other forums and yet name Mumsnet in that presentation shot on the other thread. This definitely needs to be followed up.

I'm guessing the answer is: misogyny.

I have a proposal for a PhD thesis at the University of Mumsnet (slightly better than other universities because if you spout total bollocks you will get peer reviewed and the bollocks pointed out by sane forum users, also cheaper).

An analysis of the Aston University appropriation/ stealing / data scraping of women's words. Are Aston university research staff allowed to call hundreds or thousands of women and mothers bigots for knowing about scientific reality? Analysis of whether and why data taken from a forum for predominantly women (a sex class, which is a protected characteristic) is treated differently to all other datasets held at Aston. Further analysis of whether the apparently summary judgement of hundreds of forum users (by applying blanket terms such as 'transphobic' to a huge talk forum) is reasonable, within the law, and academically ethical / rigorous.

I nominate @AlisonDonut and @Boiledbeetle as my co-supervisors if they are available.

Really? Then how do they manage deletion? Or have the Twitter/X Ts&Cs changed, or do Aston have some special dispensation?

😱

No idea. Perhaps Elon Musk should be notified.

Do we know if they accessed our real names in the database or by other means though? We need to know.

Place marking on new thread, thank you. And hello Confused Analysers. I’ll be name-changing soon which I’m really cross about, I’m attached to this one because it references one of my very favourite female authors who is always patronised, misrepresented and belittled despite making thousands of (mainly female) readers laugh and cry.

That’s one of the things about this data - it’s so valuable because it’s very very real. The trauma threads, the bereavement ones, the relationship ones, the addiction ones. The employment tribunal ones. This is what is gold to researchers because they think they can avoid the problem of the researcher impacting on the subject because the subject doesn’t know they are there. That is precisely why I think they haven’t even tried in this case to obtain informed consent, despite (I’d bet) that being a specific factor written into their university’s research ethics policies.

Plus of course, even while mining hundreds of comments posted by many different women (mostly), they still haven’t actually listened to anything we’ve said.

The change of OP on this continuation thread got me wondering about assumptions the AI and humans make about users.

The majority of time a second thread is started by the original thread OP. BUT that, as here shows, isn't always the case. So as the usual is that continuation threads are all by the same OP will the AI now assume that @VitoCorleoneOfMNMafia and @IwantToRetire are actually the same person (once they've got their heads around the fact that we can change username here regularly)?

Sounds like the sort of thing I'd like to read, could you give details. I actually thought you were referencing Cabin Pressure.

I've applied for a ticket to this, but happy to hand it over to someone who is able to go in person.

See this paper on "hate emojis"

arxiv.org/pdf/2108.05921.pdf

Do we know if they accessed our real names in the database or by other means though? We need to know.

What concerns me is that I have saved a thread from 2018 that a certain well known tra (why he put his real name I have no idea) randomly bumped last year, even though it was 1000 messages.

It was a petition and users were posting signed presumably after doing so.

An AI could certainly attempt to cross reference names on the petition with user names.

I dread to think what all the Nick Caves are going to do to the data sets.

There are so very many of us!

RethinkingLife · Today 08:48

I often want to use blatherskite but I doubt the audience for my reports or reviews would appreciate it.

There are variations - bletherskite & blagskite. When I'm on a ship, if someone says "blagskite," or more likely "the jakes," (toilet) I know immediately that they've been in the Navy.

Oi! You're not Nick Cave.

I'm Nick Cave!

Don't be stupid! You can't be Nick Cave! For I am

Nick Cave!

I've been wondering lately whether one of the ways AI will fall down will be in relation to female communication - because most of the people involved in developing AI will be men or women in male fields who normally morph themselves to fit the male dominated environment they're in (female academics tend to do this also in my experience - whether they're aware of it or not, unconscious bias innit).

All the amazing random stuff on here and in jokes, that they'll either misinterpret or just not understand.

Anyway, feels fitting to end this post with on a food note. Does anyone really like Weetabix banana mush?

😍

And the found poetry of 'Laughter in Police Interviews'

This is a good point.

It's all quite horribly dystopian.