Innovators, Public Interest Groups & Open Access Supporters Pull Out Of Talks On EU Copyright In Protest

from the squandered-opportunities dept

Back in February we reported on a worrying attempt by the European Commission to reframe the discussion about modernizing copyright in Europe purely in terms of licensing, reflected in the name of the initiative, "Licences for Europe". Although originally a series of discussions were promised to "explore the potential and limits of innovative licensing and technological solutions in making EU copyright law and practice fit for the digital age," in practice moderators shut down discussions of things like exceptions or even Creative Commons licensing. As far as the Commission was concerned, it seemed the answer to updating copyright for the modern age was just old-style licensing and nothing else.

Those events took place in one of the four "discussion" groups, covering user-generated content (although there weren't many representatives of the users present -- 75% of the participants turned out to be from the copyright industry.) However, it seems that things are equally dire in the group supposedly coming up with solutions for text and data mining (TDM), a technique whereby new information is gleaned by analyzing and comparing large quantities of pre-existing digital texts. Again, the problem was the licensing straitjacket that the Commission insisted on imposing on the discussions.

This inflexibility led to a group of eminent signatories drawn from Nobel prize winners, technology SMEs, research councils, university associations, learned academies, publishers, libraries and law academics from across Europe to send a letter to the European Commission outlining some of the key problems with that approach (pdf). Among other things, it pointed out:

Organisations in the modern world access the internet, or buy and subscribe to digital content in order for humans and computers to read and develop new ideas from what they have lawfully read. We do not accept that the right for computers to read, extract facts and formulate new ideas is a separate right to the right of humans to freely perform the exact same activities.

This is a crucial point: by insisting on licensing as the only solution to text and data mining, the publishers are trying to carve out another monopoly right, and to set themselves up as gatekeepers over the knowledge contained within the articles they publish, as well as controlling access to the articles themselves. Clearly, that places unacceptable burdens on academics, and will throttle research and innovation in Europe, simply for the sake of being able to extract a further monopoly rent for the work carried out by others.

But the European Commission paid no heed, and insisted that licensing was the only way forward for handling text and data mining in Europe. Now a group of researchers, SMEs, civil society groups and open access publishers who were participating in the text and data mining talks, have had enough, as their recent letter makes clear (pdf):

We believe that any meaningful engagement on the legal framework within which data driven innovation exists must, as a point of centrality, address the issue of limitations and exceptions. Having placed licensing as the central pillar of the discussion, the "Licences for Europe" Working Group has not made this focused evaluation possible. Instead, the dialogue on limitations and exceptions is only taking place through the refracted lens of licensing. This incorrectly presupposes that additional relicensing of already licensed content (i.e. double licensing) -- and by implication also licensing of the open internet -- is the solution to the rapid adoption of TDM technology.

This approach also undermines the considerable work that has been done in Europe to increase the amount of Open Access content available and encourage its exploitation. We are concerned, therefore, that our participation in a discussion that focuses primarily on proprietary licenses could be used to imply that our sectors accept the notion of double licensing of as a solution. It is not. We firmly believe that "the right to read is the right to mine".

The European Commission's refusal even to allow people to discuss that issue left only one option:

Given the above, and the fact that we need to prioritise our limited resources in a way which will best help the Commission to create an appropriate legal and environmental framework for data-driven innovation within the EU, we believe our contribution will be more productive outside the "Licences for Europe" framework. Therefore, we can no longer participate in the "Licences for Europe" process.

This is a real slap in the face for the European Commission, and for those who have organized these discussions in such a narrow and uncompromising way. As we saw with the user-generated content track, the sessions there were stacked with industry representatives that supported the Commission's line, thus depriving its outcome of any kind of legitimacy. With the withdrawal of such an important group of stakeholders from the text and data mining discussions, the same is now true for this key area.

As a result, the entire Licences for Europe strategy lies in tatters, since it consists almost entirely of the usual copyright maximalists talking amongst themselves, studiously ignoring the changes brought about by the Internet, and simply demanding even more while offering nothing of value. What's disappointing is that in the wake of ACTA's defeat last year it seemed that the European Commission had finally realized that it could not simply keep applying the upward copyright ratchet, and had recognized that things needed to change if it wanted to avoid more people taking to the streets in protest. The Licences for Europe fiasco shows that the Commission has in fact learned nothing.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Filed Under: copyright, copyright reform, data mining, eu, licensing, text

Data Mining Exec Pays For Burgers In Cash To Keep His Insurance Company From Knowing His Bad Diet Habits

from the privacy?-what's-that? dept

The Economist recently had an interesting article on how insurance companies are increasingly using data mining to "analyze risk." That is, they look through the data which was originally collected for the purpose of better marketing, and use it as a tool to see if you lead an unhealthy life. However, the really interesting point is highlighted by Kashmir Hill, where an exec at a datamining company admits that he's changed his habits because of this. Not his eating habits, mind you. But how he purchases food:

Insurers' interest in data mining will only grow, says Kevin Pledge, the boss of Insight Decision Solutions, an underwriting-technology consultancy based near Toronto.... Insurance firms will also analyse grocery purchases for clues about policyholders, he predicts. But that raises some sticky questions about privacy. Mr Pledge himself has begun to forgo his supermarket loyalty-card discount on junk food and pay for his burgers in cash. Promising as data mining is, much will depend on how regulators, and consumers, react.

He also notes that he's working on a system that will go through your social media profiles to see if you provide any info insurers may want to know about. This, of course, is the natural extension of our data explosion. But, in my experience, all of these companies who collect data seem to do an incredibly bad job at getting it right. And... as long as there's that "cash hole" for data, it's hard to see how accurate such information would really be.

Filed Under: data mining, health, insurance, social media

Google Books Data Mining Reveals Mad Men's Big Historical Flaw: Business Lingo

from the keep-a-low-profile dept

The TV show Mad Men has quite a reputation for going to great lengths to be as authentic as possible. The clothes, the props, the scenarios are all supposedly thought out in great detail. While some who were actually in the business at the time quibble with certain aspects of the show, it cannot be denied that the show's producers certainly go way beyond other period pieces to try to make keep everything accurate for the time period. However, it turns out that there's one area where it appears the writers have completely flopped: period-specific language. On The Media ran an absolutely fascinating clip about a researcher who has shown how frequently Mad Men uses words or phrases that were not in popular usage at the time, but only came into the lexicon at a later date:
This is actually a cross-broadcast of another podcast, Lexicon Valley, and it's covering the work of Ben Schmidt, who has produced a software algorithm that compares the Mad Men scripts... to a searchable database of language from Google's book scanning project. Schmidt's algorithm compares the language from the show with scanned books from the same period. Schmidt has a website, Prochronism, which covers his findings. I can't quite explain why, but it's really quite fascinating.

Schmidt has found that the show is pretty good about getting language about technology right (with one exception). It knows that there aren't fax machines and computers and stuff. The one area where it gets things wrong, is with the phone. For example, using the phrase "on hold." He notes that phones had hold buttons, but there wasn't yet a concept of the state of being "on hold." That showed up in the 70s.

What Schmidt has also found is that the show is absolutely terrible about getting "business" terms correct in a period specific way. That same post about "on hold" also chides the show for using "defining moment," another phrase that showed up in the 70s, but was basically stuck in academia until the late 80s or early 90s when it became a popular phrase.


Honestly, Ben's site is really fascinating. I could spend hours on it (and actually had to stop going through it post by post to finish this post). There are also discussions on phrases like "focus groups" and "leverage." But one more awesome chart from Ben, discussing the use of both "moral high ground" and "consumerism," both of which were barely in use until much later:


On the podcast, they discuss how part of the reason that the show gets the language about technology right, but not business, is because we know that technology rapidly evolves and we're more attuned to it. But people don't pay nearly as much attention to how business changes and especially how the language of business changes over time. I guess that's true, though it doesn't surprise me that "consumerism" and "moral high ground" are both more recent phenomena. "Defining moment" and "on hold" are a bit more surprising to me.

Either way, I also wanted to highlight something else about all of this that I find fascinating. For all the talk by some about just how evil Google's book scanning project is, this kind of effort and research wouldn't be possible without large scale scanning of books. While this particular example may appear (on its face) to be a frivolous (even if it's fascinating) area of research, it does highlight just how collection of certain data can open up vast arrays of data that can be mined in useful ways. When people freak out about new technologies and services, they almost always focus on how it impacts the old offerings. So most of the talk was about book scanning and its impact on book sales. But what almost no one talks about is how it enables new things that simply weren't possible before -- such as being able to build an algorithm like the one Ben built. Those kinds of innovations -- the unexpected "externalities" of projects like the Google book scanning project -- shouldn't be ignored, because there's tremendous value that can come out of them.

Filed Under: ben schmidt, data mining, google books, history, language, mad men, television

DailyDirt: Creepy Ads From Big Data

from the urls-we-dig-up dept

Lots of advertisers are turning to data mining techniques to try to squeeze more value out of their budgets. Given all the data that gets collected by our phones/browsers/credit cards/etc, it's not too surprising that ads can get pretty creepy, pretty fast. Here are just a few stories about ads that aren't technically doing anything wrong -- but that haven't quite gotten their privacy behavior right either.

• Target has been highlighted for its uncanny ability to predict when women shoppers are pregnant. Public birth records just aren't updated anywhere near fast enough for retailers who want to know when to start sending targeted ads to new parents ASAP. [url]
• Marissa Mayer said credit card companies can predict a divorce with 98% accuracy two years before it happens. Considering 50% of marriages end in divorce anyway, that might not be considered impressive..? [url]
• Facebook uses photos from some of its users to help promote various products, and sometimes the results are far from flattering. Becoming the new spokesperson for 55 gallon tubs of lubricant probably isn't what Nick Bergus wanted to be. [url]
• To discover more interesting advertising-related content, check out what's floating around on StumbleUpon. [url]  

By the way, StumbleUpon can recommend some good Techdirt articles, too.

Filed Under: ads, data mining, marissa mayer, predictions, privacy
Companies: facebook, target

UK Publishers Moan About Content Mining's Possible Problems; Dismiss Other Countries' Actual Experience

from the why-bother-looking-at-the-evidence? dept

One of the recommendations made by the Hargreaves Review in the UK was that a text- and data-mining exception to copyright should be created, with the following explanation of why that made sense (PDF):

We therefore recommend below that the Government should press at EU level for the introduction of an exception allowing uses of a work enabled by technology which do not directly trade on the underlying creative and expressive purpose of the work (this has been referred to as “non-consumptive” use). The idea is to encompass the uses of copyright works where copying is really only carried out as part of the way the technology works. For instance, in data mining or search engine indexing, copies need to be created for the computer to be able to analyse; the technology provides a substitute for someone reading all the documents. This is not about overriding the aim of copyright – these uses do not compete with the normal exploitation of the work itself – indeed, they may facilitate it. Nor is copyright intended to restrict use of facts. That these new uses happen to fall within the scope of copyright regulation is essentially a side effect of how copyright has been defined, rather than being directly relevant to what copyright is supposed to protect.

Who could possibly object to that? Certainly not the UK government, which accepted the recommendation (PDF):

The Government will therefore bring forward proposals in autumn 2011 for a substantial opening up of the UK’s copyright exceptions regime on this basis. This will include proposals for a limited private copying exception; to widen the exception for noncommercial research, which should also cover both text- and data-mining to the extent permissible under EU law.

Nonetheless, the UK Publishers Association, which describes its "core service" as "representation and lobbying, around copyright, rights and other matters relevant to our members, who represent roughly 80 per cent of the industry by turnover", is unhappy. Here's Richard Mollet, the Association's CEO, explaining why it is against the idea of such a text-mining exception:

If publishers lost the ability to manage access to allow content mining, three things would happen. First, the platforms would collapse under the technological weight of crawler-bots. Some technical specialists liken the effects to a denial-of-service attack; others say it would be analogous to a broadband connection being diminished by competing use. Those who are already working in partnership on data mining routinely ask searchers to “throttle back” at certain times to prevent such overloads from occurring. Such requests would be impossible to make if no-one had to ask permission in the first place.

Large-scale academic content mining is a pretty new and specialized field, so it's hardly likely that there is going to be a sudden mass attack of crawler-bots taking down sites. Publishers would have ample time to expand their infrastructure to handle demand as it developed, which would be to their advantage: the more their holdings are mined, the more they are likely to be cited and read. And if content mining did take off suddenly, that would suggest there is a huge pent-up demand that the current system of licensing has stifled - one more reason why it should be abolished.

Then there is the commercial risk. It is all very well allowing a researcher to access and copy content to mine if they are, indeed, a researcher. But what if they are not? What if their intention is to copy the work for a directly competing-use; what if they have the intention of copying the work and then infringing the copyright in it? Sure they will still be breaking the law, but how do you chase after someone if you don’t know who, or where, they are? The current system of managed access allows the bona fides of miners to be checked out. An exception would make such checks impossible.

This makes no sense. Infringing uses are either easy or hard to find using search engines. If they are easy to find, they are easy to pursue. If they are hard - internal uses, for example - then even miners with "bona fides" will be able to use copyright material in exactly these ways, and the publishers won't know.

Which leads to the third risk. Britain would be placing itself at a competitive disadvantage in the European & global marketplace if it were the only country to provide such an exception (oh, except the Japanese and some Nordic countries). Why run the risk of publishing in the UK, which opens its data up to any Tom, Dick & Harry, not to mention the attendant technical and commercial risks, if there are other countries which take a more responsible attitude.

The fact that some countries are already allowing content mining ought to be a hint that the other two fears are groundless. Instead, these inconvenient facts are dismissed out of hand as if the experience of "the Japanese and some Nordic countries" somehow doesn't count for UK publishers.

But as it turns out, there's actually a simple way to allay all of Mollet's fears at a stroke. At the beginning of his post he writes:

In coming to its recommendation on content mining, the [Hargreaves] Review drew heavily on the views of various strands of academia, most of which claimed that their vital research was being hampered by the lack of such an exception. The process of requesting licences of publishers was too time-consuming, it was claimed, and so an exception would make life easier.

This confirms that the text-mining issue is only being considered in an academic context – it's about giving scholars the ability to extract extra information from academic articles by performing analyses on their texts.

Now, most of that academic research is funded by the public through government grants to educational institutions and researchers, both in the UK and elsewhere. The open access movement has been pointing out for a decade that it would therefore not be unreasonable if the general public had free online access to the results of all this research it paid for - the articles published in academic journals. It would also allow many more scholars to access such publicly-funded work – including those who wanted to carry out text mining.

This would answer Mollet's fear that publishers' "platforms would collapse under the technological weight of crawler-bots." Since the papers could be freely downloaded from any one of the servers holding copies around the Internet, and then analysed on the researcher's own machine, there would be no crawler-bots involved at all. Open access would also eliminate the commercial risk: after all, what's the point in pirating material that is already freely available?

As for that competitive disadvantage Mollet is worried about, moving their academic titles to open access would actually give UK publishers a big advantage, since open access continues to sweep through the academic sector. It would mean that UK publishers were leading the way, rather than dragging their heels at the back.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

Filed Under: copyright, data mining, education, exceptions, uk

Amazon's Silk Browser To Be A Data Mining Jackpot

from the is-there-an-encryption-app? dept

We've already discussed whether Amazon's Kindle Fire 'Silk' browser is a copyright lawsuit waiting to happen, for the way it apparently is going to cache and modify content from its own AWS servers. It seems that people are realizing some other potential issues with it as well. Keith Dawson points us to a couple of interesting stories looking at the data mining implications. The first, by Chris Espinosa, summarizes the issue succinctly:

The “split browser” notion is that Amazon will use its EC2 back end to pre-cache user web browsing, using its fat back-end pipes to grab all the web content at once so the lightweight Fire-based browser has to only download one simple stream from Amazon’s servers. But what this means is that Amazon will capture and control every Web transaction performed by Fire users. Every page they see, every link they follow, every click they make, every ad they see is going to be intermediated by one of the largest server farms on the planet. People who cringe at the data-mining implications of the Facebook Timeline ought to be just floored by the magnitude of Amazon’s opportunity here. Amazon now has what every storefront lusts for: the knowledge of what other stores your customers are shopping in and what prices they’re being offered there. What’s more, Amazon is getting this not by expensive, proactive scraping the Web, like Google has to do; they’re getting it passively by offering a simple caching service, and letting Fire users do the hard work of crawling the Web. In essence the Fire user base is Amazon’s Mechanical Turk, scraping the Web for free and providing Amazon with the most valuable cache of user behavior in existence.

So there's that. Of course, there are a few caveats here. In theory, your ISP could have much of the same info -- though you can get around it with encryption. Furthermore, your ISP isn't caching everything, so there are some limitations there. Of course, on the flip side, the Silk browser is only on the Kindle Fire, meaning Amazon is only getting such data based on what people surf over that particular tablet device and its 7-inch screen. It might be interesting to see some data here, but I'd bet the sort of browsing done on such a device is not representative of how people surf the wider web.

Of course, things could get more interesting if Amazon decides to enter the browser wars in a bigger way... such as by releasing Silk as a desktop browser as well. That theory is posited in another article highlighted by Dawson, this one by Joe Brockmeier, which delves deeply into the implications of the Silk browser on a variety of fronts. He's the first person I've seen bring up the idea of Amazon entering the larger desktop browser wars:

But does it seem likely that Amazon will put that much emphasis on Silk just for the Fire? I don't think that's likely. Amazon has several jobs posted for Silk engineers, and while mobile is mentioned, it's not exclusive. I strongly suspect that Amazon is going to be releasing a Silk desktop browser eventually. Probably not in the near future – Amazon needs to make sure that its infrastructure can handle the onslaught of all the Kindle users before trying to scale to an unknown number of desktop users.

Remember when people said the browser wars were over and Microsoft had won? Yeah. Anyway, Brockmeier's article also notes that Amazon won't just have all this aggregate info, but in theory can tie the info specifically to the Amazon account holder:

Each Kindle is tied to an Amazon ID, which gives Amazon a great deal of information about you already. Introducing Silk into the mix and Amazon is going to be in a position to know a great deal about your Web browsing habits along with your buying habits and media habits. Now Amazon is in a position to know what books you buy, what shows you watch, the Web sites you visit and much more.

Perhaps it's not just copyright lawsuits that Amazon will be facing with Silk.

Filed Under: browser, data mining, kindle, kindle fire, silk
Companies: amazon

DailyDirt: Ubiquitous Sensors Are Getting Kinda Useful

from the urls-we-dig-up dept

Sensors are everywhere, recording all sorts of activities and creating an enormous amount of data. The ability to store and analyze immense amounts of information is making these sensors even more useful. Before this computational capacity was so readily available, researchers were forced to hone their hypotheses before conducting experiments. But now, it's possible to just collect a lot of data and then try to see if any hypotheses are supported by already-gathered evidence. Here are some quick links on sensors and sensor data.

• Collecting data from a wide variety of sensor feeds could become a useful service -- if only there were a standard format for sensor data and all that data turned out not to be junk. Massive amounts of garbage in, garbage out.... [url]
• Ground-based lasers could be used to map the Earth's magnetic field by exciting sodium atoms in the atmosphere and watching them fluoresce with telescopes. And this technique could be used to find underground oil deposits without using satellites to map fluctuations in the Earth's magnetic field. [url]
• British and Russian scientists should probably take a look at the previous link before they launch two satellites that measure the Earth's magnetic field fluctuations from space in order to anticipate earthquakes. Predicting earthquakes would be an awesome ability, but it's not certain there's even a correlation between earthquakes and changes the Earth's magnetic field. [url]
• You could contribute towards the deployment of 600 Geiger Counter devices in Japan -- if you donate to a Kickstarter project that aims to collect and publish the radiation data. The raw data will be published under Creative Commons 0 license. [url]
• To discover more interesting science-related stuff, check out what's currently floating around the StumbleUpon universe. [url]  

By the way, StumbleUpon can recommend some good Techdirt articles, too.

Filed Under: data, data mining, earthquakes, lasers, magnetic field, sensor
Companies: kickstarter

NY Times Finds An 8-Year-Old On TSA Flight Watch List

from the good-data-mining dept

Back in 2007, the TSA wrote a "mythbuster" blog post on its site claiming that no 8-year-olds were on the TSA "no fly" list, which apparently was a challenge to the NY Times. While it didn't actually find an 8-year-old on the "no fly" list, it did find one on the "watch list." To be fair, the TSA in its blog post does acknowledge that there could be some 8-year-olds on the watch list, but none on the no fly list. And the Times did not find one on the no fly list -- so the TSA's blog post is technically accurate.

That said, the TSA blog post also claims that if there are mistakes that lead to an 8-year-old being on the watch list, airlines can "automatically deselect" them. And... that did not appear to happen. The kid that the NYT found apparently gets vigorously searched every time he flies. The Times did get the TSA to say on record that "there are no children on the no-fly or selectee lists," but the fact that little Mikey Hicks gets searched so much highlights the problem. Right now, it seems like everything is based on a name alone. That's not particularly sophisticated. By this point, shouldn't the TSA have better tools than just a name to determine if someone is worth additional scrutiny?

Filed Under: 8 year olds, data mining, no fly list, tsa, watch list
Companies: tsa

US Government Admits: Data Mining For Terrorists Doesn't Work

from the oops dept

Over the last few years, you probably have noticed multiple attempts by state and federal government agencies to collect more data on people and to share that data more widely. This is troubling for a variety of reasons, often having to do with privacy and questions concerning the quality of data, but those concerns are almost always pushed aside by claims that this data is necessary for fighting terrorism. There's this myth out there that if we just had more data, somehow we'd be able to stop terrorists. The problem is that this is untrue. More data often makes it even harder to find the important data, and now a detailed government report has basically confirmed that data mining doesn't help in finding terrorists.

The report was put together by a bunch of well-respected academics, industry insiders and law enforcement officials on behalf of the National Research Council, and finds that the idea of plugging all this data into a machine and popping out terrorist identities is a myth. The report then lists out a variety of recommendations, concerning the use of any data mining systems, and how to make sure that individual privacy is protected. One key finding that will never be implemented: when using any such system, systematic reviews should be done to see if the systems are actually effective. If there's one thing we've seen with government programs over the years, such attempts to actually review something they've put in place to see if it actually worked are almost never done.

Filed Under: data mining, government, research, terrorism