19 November 2011

Maintaining your OLI - the problem

Earlier, I argued that online learner identities (OLIs) are pivotal in a learning ecology that is personal in that it takes the learner as its starting point and is social in that it puts this starting point at the centre of an online network of people with kindred interests (see more on this Berlanga and Sloep, 2011, Towards a Digital Learner Identity). Such an ecology thrives on the services with which it is populated. Such services come in different kinds, but these are the main categories:

• social services - they are services that intelligently match a learner with others in his or her social network; other learners come in a variety of roles, such as fellow learner, team buddy, coach, mentor, tutor, supporter, supervisor, assessor, etc., basically all the different roles teachers in ordinary formal education adopt, and a few more

• content services - they are services that match a learner's learning objectives or needs with content that could help fulfill those needs; such content will often be in the form of (preferably openly accessible) documents (explicit knowledge), but could also be in the form of implicit knowledge, only accessible by approaching the people who bear this knowledge.

The people in your social network are good candidates to fulfill the various roles in your learning ecology. And your search behaviour speaks to the things you want to learn as do, say, your blogs and wikipedia entries; they also reveal your level of expertise. Presumably, the more detailed the data about your network, about your search behaviour, your posts, tweets, etc., that is, the richer the description of your OLI, the better the social and learning services would be able to facilitate your learning. So, learning benefits from a rich OLI description.

Providing such a rich description, however, poses a privacy risk. The risk may be as grave as to result in identity theft, that is,  in somebody intentionally posing as some other person whose personal data have been stolen with the intention to harm that individual; or the risk may be moderate as when two similar but different individuals accidentally, without harmful intentions become mixed up. So the individual learner is faced with a dilemma. She should reveal everything about herself as this improves the learning experience, but she should reveal nothing at all to lower the risks involved with privacy loss. How can this dilemma be tackled? The answer is that a learner should be able to provide differential access rights to her OLI data: different groups of people get different rights. Thus, people whom one has grown to trust are provided with more rights that complete strangers. Also perhaps, people affiliated with a well-known educational institution are endowed with more rights. Etc. In this conception, controlling one's privacy is equivalent to controlling the access rights to one's data. In a next installment I will explain a schema for how this could in principle be achieved technically. However, and this is the topic of the present post, implementing any such solution which puts a user in control of her OLI data, is hard if not impossible to achieve in the current social web.

First, in the current social web data are provided freely. Web users provide them in exchange for the services that social web sites provide. So Google allows people to carry out searches, in return for the searcher's consent to Google to collect and compile a user profile, which furthers their commercial interests. And something similar goes for Facebook, Twitter, etc.  Although in principle you may decide not to agree with such schemes, in practice this is no more an option than disconnecting yourself from the electricity grid. If you want to search, you use Google, if you want to make online friends, you use Facebook, if you want to microblog, you use Twitter; etc. Second, the data are fragmented as they are scattered over various sites. This nature makes controlling them harder as you need to visit multiple sites. Moreover, sites such as Google, Facebook, etc. are walled gardens, they do not let your data escape, again because those data are the very foundation upon which their business rests. So, they are not just fragmented but your data are also deliberately kept out of your control. Clearly, in the face of this, no individual person really stands much of a chance to control his or her personal data, that is, ultimately also his or her privacy (see my earlier post on this issue).

Interestingly, even scarily if you think about it, the issue of privacy does not seem to bother the majority of the Internet users. The discussion on privacy occasionally flares up, for instance when privacy settings turn out to reveal more than previously as a consequence of a license update (Facebook) or when location data on private wifi networks turn out to have been collected and stored (Google). But the big picture of the massive amounts of data that already have been collected and stored, are used on a regular basis, fails to upset people. Wrongly so, as I have argued.

[adapted and updated December 29, 2011

13 November 2011

Why we need the Internet to stay a Commons

Thanks to Sir Tim Berners-Lee's vision we have the hyperlinked online network that we call the Web, in which we share information and ever more engage socially with each other. Thanks to the American Department of Defence the Web runs on a distributed logical infrastructure, which gives it the character of a commons: something we use and own collectively, without a central authority to govern or dictate what goes on. This character has been under threat from two sides. The Media Industries who see their business model of 'selling culture in containers' jeopardized. They focus their grievances on the peer-to-peer networks in which content is freely shared. However, independent artists of all kinds who sell their creative products directly to their fans and customers, will in the long run be even more of a threat as they cut out the middleman that the Media Industry is. This is the real game changing event as for the first time it is feasible to also cater for the long tale of people's interests. The second threat comes from governments, which are naturally inclined to fear loss of control and use national security arguments to clamp down on an Open Net. The recent measures that various governments took to make life hard for the WikiLeaks site bear witness to this.

For good measure, I should rapidly add that existing intellectual property rights need to be respected, also on the Internet; and attempts to overthrow governments, threaten its institutions or plan terrorist attacks should be nipped in the buds, anywhere, so also on the Internet. However, this is not an all-or-nothing argument. Rather, of each measure taken benefits and drawbacks should be weighted against each other. Thus, enforcing the old model of selling music, films, books in containers that one pays for or attempting to impart such a model on Internet transactions (Digital Rights Management!) stifles innovation. It keeps transaction costs high and it disallows artists and independent small producers and publishers their seat at the table. Also, imparting too much government control on the Internet brings in its wake the Kafkaesque dangers of intransparant data aggregation, data exclusion and data distortion I discussed in my previous post.

Arguments more detailed than the ones I have given can be found in a recent paper by Yochai Benkler entitled WikiLeaks and the protect-ip Act: A New Public-Private Threat to the Internet Commons He also reveals how governments and the Media Industry indeed have joined forces in their attack on the openness of the Internet. According to him, an Open Internet, one which embraces the idea of a commons is not faultless but it offers numerous and large benefits. One of those is its support for democracy and freedom: a democracy only thrives if the populace is well educated and divergent opinions are allowed to be aired; another benefit is its support for innovation and welfare: creativity thrives in heterogeneous environments, where many people gather freely and talk openly, with whom they like, when they like. In a recent 15 minutes interview Benkler reiterates this all very succinctly and eloquently.

Finally, Internet as a Commons is also crucial for the innovation of education. In a world that needs people to be better educated and needs more of them, it is imperative to experiment with different models of learning and teaching. Even though there will always be room for formal learning as in schools and universities, this cannot be the whole story (see Tony Bates' recent blog on this). Experiments with forms of informal learning or combinations of both formal and informal learning are badly needed (see also this recent report on the Future of Learning). To the extent that these are networked - and I have argued in many blogs and papers they should be - only the Internet as a Commons offers enough room for experimentation. Open Educational Resources are a key element but run of course counter to the interests of the Media Industries. Individuals as the sole owners of their profiling data are essential (see my previous post), but runs counter to the interest of governments (who want privileged access) and the Social Media Industry (who want ownership themselves or at least give people a hard time themselves to exert ownership). Long-tail education offers unprecedented opportunities for personalization and customization, but only thrives if providers of such educational opportunities have few hurdles to take, that is, on a web that is as little regulated as feasible. The easiest and ultimately most rewarding way to do this, I believe, is to defend the current character of the Internet as a Commons, surely against attacks such as described by Benkler in his paper and interview.