Computers taking archaeology jobs away

[pdurdin]

Okay, title is tongue-in-cheek. Have I missed discussion on this, or did it pass under the radar?

https://www.stfc.ac.uk/3557.aspx

[Jack]

Sounds good, though I question their implied basic asumption that we can somehow know where previously unknown archaeological sites are?? Though I agree once can map at what depths previously excavated remains are at, how deep overlying hillwash/sedimentation is likely to be and how deep modern damage penetrates. Still wont tell you where that nationally important Neolithic town is hiding.

"there are few locations in Britain where an absence of archaeology on a Sites and Monuments Record (SMR) represents an actual absence of archaeological remains" (Powlesland 2003, 275)

[Dinosaur]

It's just going to make that 'surprise' Roman town even more painful for the overconfident and poorly advised construction company - but I'm not allowed to comment }

[Marc Berger]

Stick absence of evidence up your algorithm.

Aren't they basically liars? For it to be "big" data they would have had to fed my context sheets into their ponderous algorithm and apart from the fact that most of my sheets are illegible and that I have not archived them, there is the question about copyrights (don't bother worrying about my illness). At best they have got some meta data and found a way to make their big computer map light up with a limited pallet range. I imagine that what they are trying to confuse everybody with is that the big BGS map is the big data. Thing is if you look at BGS sample density it has not much to write home about. Can anybody be bothered to wonder what their predictive algorithm does with the terms "there be badgers" and "probable post hole". I always took it as read that every site must have these and that they are only reported to Defra and the heritage nannies if they are not there but when we still reserve the right use the word "possible" if questioned in a court of law. Apart from a very sad webpage arent these people refuges from the email tapping world. Their contacts web page leads to this claim to fame, note the use of "worked". This appears to have been the CEO. https://plus.google.com/101475383433139483881/about

Is Democrata dead? Long live the Republic.

[GnomeKing]

Stick absence of evidence up your algorithm......Aren't they basically liars? .

lol

Pretty much - what a load of over-engineered cobblers - to borrow the military term - ' Mark 1 Eye-ball wins every time'.....or should that be trowel?

[Dinosaur]

Unfortunately, some applications of 'Big Data' do work (e.g. American police forces use it quite successfully to work out where/what crimes are likely to be committed on a day-to-day basis, down to particular neighbourhoods/streets), but I can't imagine how it could work in Archaeology

[Marc Berger]

The American police presumably have not quite successfully yet big dataed their shooting of innocent people.

PS I claim big dataed in this context as a new phrase and spelling.

[Land-Escapist]

It's all very well and good but the trouble is with this archaeology is that it was created by humans. And humans have the annoying thing of being so damned unpredictable with their activities and lives and dying and such... So whilst you can to some degree 'predict' where activity will have happened, at the same time you really cannot.

[barkingdigger]

Since computers can't think, their algorithm-generated predictions are only as good as the data they have to chew on! There be Planning challenges in this Brave New World.

Anyone remember when the map of Roman Britain bore an uncanny resemblance to the motorway network?

[BAJR]

You can automate data collection, but not interpretation.