Monthly Archives: January 2005

bender changing

My machine bender (swordfish.rdfweb.org, sw1.ilrt.org) is getting a new brain and so my photos, demos and such will be down for up to a week from Tuesday. Normal service will be resumed as soon as possible.

FOAF and ID cards

In the UK, compulsory ID cards and a national register appear to be a near-certainty in the next few years. To many people in the UK it seems an obvious and sensible reaction to problems of benefit fraud, terrorism, asylum seekers and certain sorts of crime. To me it looks much more like a partial solution to a data integration problem: an ID number will be a unique key into the multiple databases held by government for a given individual. The police force in the UK has a geographically distributed structure, and a simple way of identifying individuals should help them manage data across these organisations more effectively. Likewise, whenever the data is shared across government departments, a unique key for each individual should make sharing data simpler, reducing costs in the long term and perhaps, reducing errors.
In FOAF, in a small way we’ve touched on many of these issues of data merging, data integrity and trustworthiness, and on some of the implications for privacy. A strong motivation for FOAF was “I want my data back” – from the organisations who control it whenever we register online to buy something or share some information or talk to others. Straight away we ran into the issue that in talking about ourselves we talk about others: FOAF data is about people and their relationships as well as the things that they make and do. In choosing to make available information about ourselves we quickly segue into making information about others available too: “I want my data back” isn’t that simple.
Aggregating this information can have unexpected effects. It’s hard to predict the social impact of these effects: we’ve seen positive – the personalization of news and the concomitant connectedness we can feel towards people far away in circumstances and geographically. The negative impact is in the unexpectedness of connections made and the unpredictability of who makes them. Privacy through obscurity starts to disappear. This isn’t new: data mining can suggest connections not apparent at first; but FOAF and the Semantic Web are likely to make the connections easier to find.
So what? well, two aspects of FOAF experience are relevant to ID card discussions, I think.
One is that there’s a great deal of ignorance about the effects of merging data from different sources, particularly in the wider, non-technical community. Danny Weitzner has spoken eloquently about this, but the debate is not happening in the UK at least. FOAF doesn’t provide any answers but the FOAF community – like much of the online community in general – is fumbling towards the questions.
Secondly although FOAF data is open to the world while the personal information held by government is limited in its accessibility, the limits on its distribution are not clear and seem likely to be quite broad. For example the company Atos origin has multiple contracts in the UK: in the Metropolitan police force, several hospitals, the immigration and nationality directorate, and the department of work and pensions. This is just one company of many who work in outsourced parts of the government. Data management within government involves sharing information with many more organisations. Where the effects, limits and costs of data integration at this level are not understood by the populace, the checks on government and business are very limited indeed.
On a personal level, I find the notion of giving any and all government the ability and right to combine multiple pieces of information about me repellant, especially given their attrocious record on computer projects and the opinion of the Information Commissioner that ID cards will make identity fraud easier not more more difficult; and various rumours and consultations that imply that businesses and medical staff will be used as immigration officers. If the system works it will be an effective means of excluding further those who already find it difficult to get work, benefits and healthcare. If it doesn’t work it will be a huge and costly mistake. We aren’t just data to be managed for the convenience of government and business: they work for you not the other way around.
Links: ID cards bill and documentation, noID campain against ID cards and the national register, the Register, Atos Origin’s ID card white papers.