I don’t know where to start. I really don’t “text mining” oy ~ctm
Public Release: 3-Aug-2017
A new method of surveying published research, which has highlighted current issues faced by the marine environment, hopes to place scientific knowledge at the heart of policy agendas
Frontiers
Covering two-thirds of our planet, the ocean was once thought to be too big to be threatened by human activity. Scientific evidence now shows that our use and abuse of this environment is having a detrimental effect on marine habitats across the globe.
New research, published in the open-access journal Frontiers in Marine Science details current threats to our coasts and oceans, and potential solutions to these problems, as detailed by the focus of research undertaken by marine scientists over the past ten years.
“The amount of marine research being published is growing rapidly,” says Murray Rudd, an Associate Professor of Environmental Sciences, who carried out this research at the Emory University, Atlanta, USA. “There are a number of topics, such as the effects of climate change, marine plastic and debris, conservation, and increasingly, the human and social dimensions of ocean use and management, that are now the focus of major and emerging research efforts.”
Professor Rudd sifted through the tens of thousands of relevant scientific studies from the past decade using a method called ‘text mining’, which has only just become possible with recent advances in computing power and software. This method involved large-scale scans of written text, to assess how different words and phrases are used.
“I looked at the summaries of over fifty-thousand ocean and coastal science research articles and assessed how language relating to ocean challenges and their solutions varied over time. That allowed me to identify trends, emerging ‘hot topics’ in ocean research, and to see what types of solutions were most often proposed for specific challenges,” explains Professor Rudd.
As well as revealing the problems faced by our coasts and oceans, the survey highlighted how some of the solutions put forward to resolve these issues differed in focus. For example, Integrated Coastal Zone Management, a strategy that aims to ensure all the interests of those involved in the development, management and use of the coast are taken into consideration, was frequently mentioned in conjunction with sea-level rise, but few other issues. In addition, marine protected areas, once thought to be a tool for fisheries management, are now also seen as a way to help buffer ecological systems from predicted climate change effects, such as changes in ocean chemistry and temperature. The survey also found that the BP Deepwater Horizon oil spill in 2010 generated a surge in focused research relating to this catastrophic event.
Professor Rudd hopes his research will demonstrate that the method of ‘text mining’ scientific publications will enable governments and international institutions to methodically track and identify important and emerging ocean sustainability issues, and the organizations and scientists who work on them.
“Oceans and coasts are crucially important for human society, especially for the 2.4 billion people that live within 100 km of the coast,” he explains. “We have seen some issues that are very important to the public – marine plastic pollution is a good example – rapidly receiving increased research effort. We don’t fully understand however, how emerging issues of concern catch the public’s attention and get onto government’s policy agenda, nor how scientists can and should communicate their research findings so that new knowledge about ocean and coastal sustainability is put to good use. Future research on the dynamics of information exchange between scientists, the public, and decision-makers in government and business is a natural next step for some of the specific research topics that were flagged in my exploratory study.”
Professor Rudd concludes by advising we combine all our global scientific knowledge for the good of our marine environment. “To sustain the oceans, and the benefits they provide us, we need to be able to use all the information from scientific research to help decision-makers weigh-up options about how we use the ocean and where / when we need to protect it. The ocean is dynamic and complex, often with long time lags before we can reverse the consequences of human activities, so building our scientific understanding is essential, if future generations are to inherit healthy oceans.”
###
Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.
Text-mining tells you nothing except what people chose to print out of what people chose to write about the subject you chose to investigate.
I have a hard time taking this “text mining” seriously, but there may be a tiny chance there is some tiny kernel of something meaningful there.
May be of more value to sociologists studying humanity’s weird enthusiasms than to marine scientists however.
obviously the only “trends” you will detect by this method is what is “trendy” to publish. This should not be confused with any “trend” of climatic variables.
don’t forget this is a pay-to-publish journal which are usually very light on peer review, if they do any at all. The main aim is to cash in the author’s publishing fee, not to reject it.
I think data mining of articles on climate research over the past 50 years should be used by skeptics to show the increase of words that hedge — like: “might” “may” “could” etc. — and phrases like: “models show” “requires more funding” etc. — as well as the increase in the number of years before projected tipping points and catastrophes occur. — lots of good fodder here to throw back at them!
Bingo. “If a million flies eat….
Confirming confirmation bias.
So he used Google Trends to see what topics people looked up and assumed searches were indicators of actual problems?
The (open) article linked to in the following post demonstrates that the method of “text mining” is crap in that context:
United Nations has taken part in scientific malpractice – and contributed to an overly negative perception on the state of the ocean!
Here is one quote from the article:
“Duarte et al. (2015) have argued that poor citation practices are one of the elements that have perpetuated perceptions on ocean calamities (including rising jellyfish populations) that are contributing to an overly negative perception on the state of the ocean. Our study confirms that that mis-citation facilitated the perception of rising jellyfish populations.”
Cynically, what the common themes in oceanographic research seem to have in common is what the author thinks will draw funding.
+10. Lets see… ocean acidification. Dying coral reefs. Sea level rise.
And a real one … Overfishing.
DK, my own list of real ocean issues includes species overfishing, coastal pollution (e.g. This years Mississippi dead zone), and bycatch. But those will not garner AGW grant money. I left off the sarc tag thinking the irony was sufficiently obvious.
The irony was obvious. It just seemed to me appropriate to note that not all the standard list of alarmist threats are totally baseless.
Additional thought. I’m not sure the “Mississippi dead zone” is abnormal. Seems to me that the black Cretaceous and Paleozoic shales (Marcellus, Utica, Bakken, etc) being drilled for natural gas and oil probably started off ever so many million years ago as rather similar anaerobic sediments.
Another thing they won’t notice is the brilliant EU method of saving fish stocks by using quotas that mean when you’ve caught your quota of for example cod you throw back perfect;y good fish which are dead or dying.
Hopefully once were’re free of the EU those with some idea of how to really look after fish stocks will be allowed to have a say in new rules.
I’m not going to hold my breath though.
James Bull
May as well poke fun at this sort of nonsense. No one seems to notice they aren’t paying attention to “poking fun”.
These folks take themselves very seriously and they really don’t care what you think. By now it should be obvious this isn’t a debate. There’s nothing scientific about it.
You’re going to do exactly what you’re told to do. No if’s, and or buts. You all seem to think this is a democracy and you can make “rational argument”. You’re very wrong about that.
++
Yep…
Calling this research is a bit of a stretch. More like regurgitation renamed as text mining.
It isn’t research Rudd, it’s an edict. Sit down, shut up an do it.
Seems more like digital archeology than anything.
Is this like the DARPA BS detector program?
David Middleton
August 4, 2017 at 12:20 pm
“Is this like the DARPA BS detector program?”
————-
it seems very much like so. 🙂
It seems like a proposition of a very quick and fast BS detector, with a Torch attached to it, which can quickly and easy point out at big amounts of suspect BS……:)
cheers
D(ramatic) A(rticulation) of R(epetitive) A(rtifices)
Oops, left out the P—which could take many forms but for oceans this is the best one.
D(ramatic) A(rticulation) of R(epetitive) P(ellagic) A(rticifices)
DM, how do I get one of those? Got a real good volt/amp/resistance meter in my workshop, but have been searching for a good simple BS meter. Closest I came was ebook The Arts of Truth, but that is long on thought processes and requires intellectual diligence. Hardly a simple BS meter.
I got a condensed one here:
The principles of science (v7.5)
I´m always looking for scrutiny by the way – be my guest.
I know that this may be off topic but I am reminded of of the Foundation novels written by Isaac Asimov. In these books research papers were written by performing searches of all of the papers which had been written on a particular topic rather than actually performing original research. I have seen another article or two which were written using similar methods and it makes me wonder whether the author was perhaps a prophet of sorts.
I know that this article does not fall exactly into such a category but I do worry that perhaps a trend is beginning.
Also performing such a search does not allow room it would seem for determining the accuracy or validity or the work and would perhaps lead to a blind acceptance of the results of a paper. This is especially troubling considering the ease with which fake papers have been accepted for publication recently.
Sounds like Cook et al., 2013… 😉
Actually review papers are a legitimate and useful scientific genre. However they take a lot of time, konwledge and work to do, at least well, and are not highly thought of. However doing a good review papar is a very effective way to get a lot of citations.
Yes. I have done recently a lot of reading scientific literature, and though reviews sound like not real work, a good review tells what some community thinks about research topic at hand, and summarizes what the author knows. I like them, a lot, over papers which use their textual effort to explain the boring methodological means to reach a result that does not bring any broader view to the subject and cites badly existing literature revealing bad comprehension on existing data.
Anyway, both original research and reviews are necessary.
Often sifting through the original papers is time consuming and requires knowledge on how far the science had advanced at the time when the paper was written.
Agree totally. I’ve found this to be a wonderful resource.
http://www.springer.com/us/livingreviews
“Living Reviews are scientific open-access journals, publishing review articles that provide insightful surveys on the research progress and literature in the fields they cover. Its unique concept allows authors to regularly update their articles to incorporate the latest developments.”
Fields covered are Solar Physics, Computational Astrophysics and Relativity.
I remember that character. High Lord Something or Other. He made lots of promises and spouted a great deal of verbosity. They did some sort of “text mining” to get at the actual meaning and the computer finally came to the conclusion that he had said — nothing!
Isaac Asimov was a fine scientist, he was a Professor of Biochemistry, as well as a great science fiction author. So I am sure that he wrote from experience when he described the decay of science in the Foundation Trilogy.
“research papers were written by performing searches of all of the papers which had been written on a particular topic rather than actually performing original research.”
That sounds like IPPC
{T}o methodically track the evolution of the enviroprofiteers’ marketing campaign (a.k.a. “propaganda”).
Text mining is a substitute for reading and thinking. How else is a man gonna get through tens of thousands of pieces of junk science.
I think the idea looks very promising for confirming the various kinds research that scientists expect the government will fund in the future and how much money will be available for it.
What a head-shakingly obvious half-truth this is:
….the ocean was once thought to be too big to be threatened by human activity.
The ocean, in toto, still IS to big to be “threatened” (your money or your life, ocean!). Lol.
Such a “researcher” would also assert such nonsense as:
“Mt. Everest was once thought too big to be threatened by human beings. We see that it now IS threatened because humans are wearing away its surface by clambering up it and by leaving garbage that THREATENS the MOUNTAIN’S ecosystem.”
And did you know!
“The elephant was once thought too big to be threatened by ants. We see that it now IS threatened because when ants run in droves across the elephant’s pathways, they cause the elephant to suddenly stop to be sure there isn’t a deep crevice in the path which THREATENS the ELEPHANTS by wearing down the bottoms of their feet.”
Many fish species have been decimated though. They are not “the ocean” though.
The Northern Cod species have started coming back to the places they left when cyclic warming drove them north. In fact, if push comes to shove, fishers will go broke before they can extinctify species.
Associate Professor Rudd, sifted; a subjective description for a subjective process.
Another model that allows alleged associate researchers to find what they want to find.
No science involved.
” which has only just become possible with recent advances in computing power and software. ”
I don’t know where the good professor has been hiding, but people have been text mining for 30 years.
Whoever yells the loudest will be rewarded. I know of several communities living on that principle.
A google search is literally mining the strings of web page key words and page content. This is what science has come to? Really?!?!?! I spent a few years working on a system for NOAA to search indexes of geographic data sets via keywords, coordinates, etc. There is enormous opportunity in these searches to have false positives in the results and limited ability unless you actually read the content to weed out the false positives. This will produce an exaggerated result for whatever search is launched.
This is hard wiring confirmation bias into the process. Engineering it into the process. Amazing.
and scary
if you did it that way no wonder you got garbage.
Its the BEST way.
Let’s see. If you consolidate a bunch of junk science articles you end up with a bunch of JUNK.
Yep.
http://www.svilendobrev.com/kartinki/wall-e-at-work.jpg
Wall-E
(youtube)
Professor Rudd: Ta da!
Posted some before, worth repeating
An article on “The New Generation of Sea Scientist” appeared in Oceanography Magazine quite a few years ago. A few quotes—-
“Now, however, seafaring adventures are a much smaller part of the way we perceive our careers than those who are 15 or 20 years older,”
“…….of the students who joined her Ph.D. program the same year, only one, who focuses on biology, has relied on data collected on ocean cruises for their graduate research, she says. THE OTHERS HAVE USED REMOTE SENSING DATA, MODELING STUDIES, OR DATA FROM THE ARGO.” “And WHOI oceanographer Peter Wiebe is dismayed that the institute’s graduate students routinely turn down invitations to take a berth on an upcoming cruise. ”
I know a world traveler, flies a plane, upon returning from a sail trip from the Chesapeake to the Virgin Islands said that he did not know the ocean was so big and where was the plastic?
THERE IS NO OCEAN SMELL FROM A COMPUTER, YET!!!
FISHERIES, LOW OXYGEN AND CLIMATE CHANGE: HOW MUCH DO WE REALLY KNOW?
http://onlinelibrary.wiley.com/doi/10.1111/jfb.13203/full
It takes only a short time to find a new paper (an admitted minority but I have plenty of old ones) with caveats about most anything in the ocean crisis portfolio. Of course, they always need to know more. Paywalled, but one sentence from the abstract.
“Some fishes may move away from areas until oxygen levels return to acceptable levels, while others take advantage of a reduced start response in prey fishes and remain in the area to feed.”
How about that, fishes tolerating low oxygen! If a real crisis arose on their computer, they would miss it and they are ruining the credibility of my profession.
And for an old one (“The whole sea seems to die…” Clark, 1929, quote about a tilefish mortality in 1880; Collins, J. W. 1884. History of the tilefish. Report of the U. S. Commission of Fish and Fisheries. 10(1882):237-292.)
{Observed over the past 20 years or so}
— As children, such young adults didn’t care to climb trees or ride bikes.
— As teenagers, they actually PREFERRED to have Mom drive them than get their own driver’s license (! unbelievable).
— As adults, we see from the above article that they don’t like outdoor adventures and cannot write English (for a formal publication — I’m not counting blogging, lol) as well as a 4th grader can (er, could…. ).
— No wonder that for a job, these scientists who style themselves “earth” or “atmosphere” or “biology” experts or the like:
sit inside
in front of a computer
reading information gathered by others,
information which, however precise and accurate,
represents only a tiny % of the total data of earth and its atmosphere,
guessing over and over what that information might mean,
and that
is
all.
— One might call this “a touching example of the beautiful thing that is a child’s trusting naivete.” One would be more accurate if one called it what it really is: “a serious, chronic, brain dysfunction manifested by the absence of normal scientific inquisitiveness” (in addition to mere “sloth”).
Text mining is to the 21st Century what mucking out the barn was to the 19th. Except that the product of text mining cannot be used as a soil conditioner.
Like Metadata, this sounds like ‘Meta-Science’. Google’s search engine does this, data capture, trending, etc, it is nothing new. Sarcastically, when you cannot do science yourself, you can make a career out of cataloging other people’s science. I see nothing wrong with it in theory.
Ah No. Googles search engine does not do topic modelling. Nor does it do semantic coding.
Its coming though.
So you say.
Those “Decision Makers” scare me.
The big ‘white chief’ of the coal mining gives not a dime for the BS ‘text mining’.
+1! 🙂
And America (true America, the millions of us who could not care LESS about the alleged Russian “influence on” our elections) digs Trump!
Hi Ms Moore
This must be some kind of a record.
Three generations of Tyler family have lived through all of the US 45 presidencies. John Tyler was born in 1790 and was the 10th President of the United States (1841 to 1845). It appears that President Tyler still has two surviving grandsons Lyon Tyler Jr. born in 1924. Harrison Tyler born in 1928.
Can we please borrow the Don for a while ? I promise we will send him back , eventually .
Sorry, One of the biggest donors to Trump was a pioneer in Computational linguistics and his
Analytics company was instrumental in Trumps win.
Can you prove that or are you just spouting garbage, again?
Vestas offshore test turbine on fire.
http://m.nasdaq.com/article/big-mhi-vestas-test-wind-turbine-on-fire-20170804-00966
Here’s a practical application for text mining for you – look at the number of turbine fires reported in the last decade against the number of turbines deployed.
Why don’t you do that Griff, given the fact your post seems to suggest there will be “low” numbers.
So text mining provides policymakers with non issues like ocean plastics, sea level rise, coral bleaching, acidification and all the usual non starters in the scare game, but leaves out an important paper that says almost all research studies of ocean ecology problems have been very badly done. The paper by a world renown Ocean researcher was reported on here at WUWT and it even offered guidelines for doing meaningful studies of the ocean, marine life etc. Maybe someone has a quick link to this important paper. This mine’s product is a resource to serve global тоталiтаяуаи policy.
It’s good that WUWT keeps us informed of these, the latest developments in scientific insanity.
lol
How would he know whether he was mining Gold or Pyrite?
When this text mining technique is perfected, it will serve superbly as a way to cull out the 97% chaff that chokes up the scientific literature. We will also pick out the worst 10 papers to use as object lessons on how not to do a scientific study. Dr Rudd’s work likely will turn out to be a valuable case history on how to cull out the scientific slag. His buzz words for drek selection could earn him a Nobel Prize.
Dang,
Now nobody minded the technics of text mining ( basically just applying math to texts) when the
texts in question were proxy studies, and Wegman showed the social network of Mann et al.
And no one complained when I used a tool of text mining ( search for high entropy words0 to bust Gleick
And no one complained when Anthony used the tool of text mining I recommended to him to study whether
Gleick was the author of the forged memo.
There are many tools in the textual analytics tool chest. A lot more than when I started looking at it back in
the early 80s.
The author is doing basic preliminary Topic modelling.
https://www.youtube.com/watch?v=4p9MSJy761Y
here are the questions he wants to answer
” Specifically, my research questions in this text-oriented exploratory research included: Are there differences in the frequency with which keywords or phrases relating to instruments or strategies (i.e., potential solutions) co-occur with those relating to different types of ocean challenges?; Does the frequency at which text associated with particular ocean challenges and possible solutions vary temporally across all articles?; and Does the frequency at which particular text associated with particular ocean challenges and possible solutions vary between relatively highly- and lower-cited articles? The latter two questions are motivated by the idea that cutting edge ideas and methods that define and drive epistemic communities might be productive areas for further detailed assessment such as expert interviews or bibliometric research. ”
long ago we went at this by building concordances. I built one for Climategate 3 mails.
Essentially you are tring to take a huge corpus and make it more manageable, and look
for gross trends and changes. Places to dive in deep.
For example, if you searched the climategate mails you would find one occurance of the phrase
“squeaky clean” That was in the mail where Jones discussed making sure a particular paper got
a friendly reviewer.
if you search for the word megaphone, you would find one mail. It was about the creation of realclimate.org
These high entropy words ( Shannon entropy) Are clue that something interesting and important is going on.
basically something novel.
Here is the main genius
Its machine learning boys and girls, get used to it.
At one job I use this to analyze notes from field reps who visit customers.
Thanks for the summary.
Intel agencies use these techniques frequently from what I understand. And do so in ways far more focused and particular than this schlock faux study.
Search engines, network analysis, etc. all are basically spying tools.
That the climate obssessed use it to “help” prove their apocalyptic claptrap is not surprising.
That must be a cut-and-paste story. In many years reading this site I have never seen you post so much rubbish in a single post!
A friend of mine wrote a program that used text mining to automate stock picks. He conducted a hypothetical test and found, had he actually turned the program on and used real money (all things that are possible to automate with online stock purchase websites), he could have made a 200% return in less than a month. So data mining has its uses, but as a scientific vehicle…?
I live 100 meters from the ocean on Sagami Bay in Japan.
The most tragic environmental problem in this area has been the Japanese government’s approval of fisherman using ultra-fine sieve nets (2mm x 2mm) to catch baby sardines, which are used to make a local delicacy called “Shirasu” (sun-dried baby sardines sprinkled over rice, baked in bread, sprinkled on pizza, placed in omletes and salads, etc., ).
Obviously, the sardine population has crashed 90% since this totally INSANE fishing practice went into effect, as have ALL fish populations that feed on sardines, or are inadvertently caught in the ultra-fine sieve nets)
I often call the Japanese EPA and Fishery Department to complain about this deplorable fishing practice, and the feckless government hacks have the audacity to blame CLIMATE CHANGE (温暖化) for the sardine and other fish species population collapse..
Various: Fishery, Tourist, Hotel, Restaurant and Supermarket lobbies have strong sway over government hacks and spend a lot of money keeping this insane ultra-fine sieve net practice in place so they can continue harvesting and selling shirasu.
This is an example of CAGW nefariously being used as an exuse to coverup the true underlying cause of deplorable and real environmental problems.
Thank you Samurai. This is a huge problem around the world.
Here in the US, government blames “climate change” for neglected infrastructure failures.
In the Philippines “climate change” was blamed for deaths and destruction due to unenforced building codes when a Typhoon struck.
This social madness is quite lucrative.
search internet where data contains ‘climate’ and ‘serious’ and ‘man-made’ and ‘worse’ THEN publishAs(“confirmation of problem”);
Overwhelm the system with b.s..