Latest activity → Feed

June 2016

Short thought

There have been lots of really good things today. Libraries joining OLH, some great articles published, fantastic calls with good people.

Short thought

George Osborne, you know, the *Chancellor* of the UK, has been mightily quiet on a day when the country's currency fell 10%.

Short thought

The perennial week of the summer is almost upon us where, on average, there are 5 events, spread nationwide, per day, that I want to attend.

Short thought

If your library isn't yet supporting @openlibhums, why not suggest it today before the year's budget ends?

Short thought

Renewal stats for @openlibhums are great! All but four UK & US institutions renewing (and those 4 haven't said no, just not responded yet).

Short thought

Takes me a good hour to get into the mindset to work with it. I confess, it is quite elegant once you're in the right frame of mind.

Short thought

Been doing lots of XSLT work today. It just remains such a strange language to me. Driven by document rather than object- or logic- flow.

Short thought

Spent morning setting up a build environment on new machine. Constant failures. Assumed it must be me. Nope, Foundation 6's browser support!

Short thought

This referendum is a sick in-fight for the Conservative Party leadership in which virtually nobody is qualified in knowledge to vote.

Short thought

Jo Cox, MP. Worked for anti-slavery charities. Staunchly supported those refugees who are worst off in the world. Shot dead. Horrendous.

Short thought

Stacked percentage representation of PoS trigrams in chapters of Cloud Atlas that differ by 1.9 standard deviations.

Post

Consortial funding and downward price pressure for open access

Different groups of open-access advocates want different things to be achieved by OA. The "OA movement" is not a homogenous group. Some members of the group believe that all publishing labour is unncessary or could/should be volunteerist. Others want to allow people to read green open access…

Short thought

Of course, the accuracy level is off-the-scale bad for Sloosha's Crossin'... I need to exclude that.

Short thought

1,000 most-common PoS trigrams clustered with 90% consensus are enough to correctly tag all chapters.

Short thought

I used the Stanford tagger to PoS tag Cloud Atlas. Trigram PoS signal weaker for clustering different narratives but still present.

Short thought

Can confirm that Stanford POS tagger struggles with Buffalo buffalo Buffalo buffalo buffalo buffalo Buffalo buffalo.

Short thought

My love to and solidarity with all LGBT people affected there or continuing to be persecuted anywhere.

Short thought

I don't say anything after most US shootings since tweeting about them feels inane & insufficient. But this is such a horrendous hate crime.

Short thought

The Glass Bead Game is an extraordinarily good novel that I wish I'd read before now. Feels like a 60s Borges running at a meditative pace.

Short thought

The twenty most frequent terms (still common words like "he", "is", "that") provide confident discriminators for all sections using delta.

Short thought

You can distinguish between all-but-one chapters of Cloud Atlas based just on the frequency of words “the”, “a”, “I”, “to”, “of”, and “in”.

Short thought

I do say: you won't see anything but trust me, I have a raging infection. But they look so disbelieving...

Short thought

Because it's a rare drug, most clinicians have no idea of this and keep checking my temperature & c-reactive protein counts.

Short thought

Among the weirdest things re the biological immunosuppresants that I must take is that they mask fever & inflammatory markers in blood tests

Post

Initial parsing work on large JSON corpus

Yesterday, I wrote of a challenge that I faced in working out which texts in a corpus have decent OCR and, then, which texts they actually are. This morning, I put together a small script that has a first go at this. I enclose this below for anybody who is interested.

Short thought

If only my twitter feed were voting in the EU referendum, I'd feel far less nervous.

Short thought

Car: is that a person?
[PROCESSING]
Car: nah, it's "<\u00a9jcbfrrp'. ..fcitioiu". Full speed ahead.

Short thought

Does it not worry anyone else that we're making self-driving cars but that our OCR techniques still give <\u00a9jcbfrrp'. ..fcitioiu *4 JVM?

Short thought

But I have done some good writing this morning and am pleased with progress on the metadata book

Post

Identifying 26GB of JSON Novel Data

For part of one of my current research projects I have a pretty large (26GB) corpus of digitized JSON novels. I'm interested in ingesting these and then performing various methods of authorship attribution to them using delta, nearest shrunken centroid and other techniques.

Short thought

Hmm. Been waiting for a USPS international package for a couple of weeks. Today it says it was "delivered", but no item arrived...

Short thought

It will also be the first full fun outing of my SankeyTextualVariant D3.js software modifications for visualizing textual genetics.

Short thought

The full article that underpins the presentation is under submission and in what I hope is a final round of rigorous review at @openlibhums.

Short thought

I'll also be presenting the internal publishing history of the text and a stemma of the translations and film script.

Short thought

Creating my slides this evening for @WHN21CWriting where I will be presenting my forthcoming work on the version variants of Cloud Atlas.

Short thought

Almighty thunderclap scared the life out of me and seems to have knocked our entire street's broadband offline.

Short thought

Also, DeLillo claimed not to be able to remember protagonist of Cosmopolis's name (Erik Packer) in a sentence were he used the word "packed"

Short thought

Though I'm doubtful that he has forgotten as much about White Noise as he claims.

Short thought

Don DeLillo was fantastic this evening. Questions were also good - lacked the pretension that usually dogs such events.

Short thought

If the UK votes Brexit, what actually happens is that England is renamed David Attenborough & the Isle of Man goes, right?

Short thought

Would anybody like a spare ticket to Don DeLillo this evening at the Southbank?

Short thought

I always wanted to play sax as a teenager but my mum admitted that she made me play clarinet because I was the academic child (!)

Short thought

Bidding farewell to the lovely Dartington Hall, which has hosted the #DARTS5 event for the past couple of days.

Short thought

Off to Dartington tomorrow to keynote DARTS5. My wife is coming too since it's school half term & may listen to the talk. Now I'm nervous.

Short thought

Today's news brings more people who went to university for free & with grants telling students how lucky they are to be able to have fees.