Seus dados e o Facebook

Vídeo sobre o que o Facebook faz com os dados que disponibilizamos (via Panel de Control)

Databases

Hoje, mais um artigo do The Guardian divulga relatório sobre os bancos de dados públicos na Inglaterra, incluindo o banco de dados de DNA, que indexa todas as crianças inglesas. O relatório, realizado por membros da Foundation for Information Policy Research, aponta sérios problemas quanto à efetividade dos dados e quanto à privacidade dos indivíduos, tal como o banco de dados de obesidade infantil. As crianças também são objeto de predição a partir desses bancos de dados, os quais visam antecipar futuros problemas comportamentais ou de saúde. Trecho:

"The report says children in particular are placed at risk. Three of the largest databases set up to support the young are failing to achieve their aims, it says.

Terri Dowty, of Action on Rights for Children, said young people had never been so measured, graded, monitored and discussed; the level of intrusion could not be "justified on the basis of good intentions".

The report raises concerns about the Home Office system, ONSET, which gathers information from many sources to predict which children will offend. The report says children could be stigmatised by a system that contravenes the European convention on human rights."

Catalogue

Rodrigo Firmino mostrou, no evento sobre vigilância na PUC-PR, esse vídeo de Cris Oakley, que explora as conexões da vídeo-vigilância com os bancos de dados e a elaboração de perfis proativos. Catalogue:

Bancos de dados e política

Um estudo sobre as implicações políticas dos 600 bancos de dados públicos e privados que contêm informações sobre indivíduos no Reino Unido será divulgado no próximo mês pelo thintank Demos, conforme matéria do The Guardian.

Bancos de dados e vigilância

Matéria do Economist.com trata do enorme aumento dos procedimentos de coleta de dados sobre indivíduos em nossa cultura, os quais que vêm ultrapassando em quantidade, diversidade e agilidade aqueles empregados nos regimes totalitários. A matéria trata ainda da ampliação de outros sistemas de vigilância, como câmeras CCTV, etiquetas RFID etc. Trecho inicial:
"IT USED to be easy to tell whether you were in a free country or a dictatorship. In an old-time police state, the goons are everywhere, both in person and through a web of informers that penetrates every workplace, community and family. They glean whatever they can about your political views, if you are careless enough to express them in public, and your personal foibles. What they fail to pick up in the café or canteen, they learn by reading your letters or tapping your phone. The knowledge thus amassed is then stored on millions of yellowing pieces of paper, typed or handwritten; from an old-time dictator's viewpoint, exclusive access to these files is at least as powerful an instrument of fear as any torture chamber. Only when a regime falls will the files either be destroyed, or thrown open so people can see which of their friends was an informer.
These days, data about people's whereabouts, purchases, behaviour and personal lives are gathered, stored and shared on a scale that no dictator of the old school ever thought possible. Most of the time, there is nothing obviously malign about this. Governments say they need to gather data to ward off terrorism or protect public health; corporations say they do it to deliver goods and services more efficiently. But the ubiquity of electronic data-gathering and processing—and above all, its acceptance by the public—is still astonishing, even compared with a decade ago. Nor is it confined to one region or political system."

Fábricas do gosto e Redes sociais

As redes sociais na Internet, como Orkut, Myspace, Friendster etc, não apenas refletem novos modos de sociabilidade e de exposição do eu no ciberespaço, como também constituem fontes valiosas de informação que alimentam bancos de dados, agentes e sistemas de classificação/fabricação de gostos e identidades. Notamos aí um elemento importante dos regimes de monitoramento e vigilância no ciberespaço. Um interessante artigo sobre esse tema me foi indicado pelo querido Henrique Antoun: Unraveling the Taste Fabric of Social Networks.

O artigo é assinado pela Pattie Maes (em co-autoria com Liu,H. e Danenport, G.), referência importante na pesquisa em inteligência artificial e na produção de agentes informáticos de recomendação de gosto, tais como os que encontramos na amazon.com. Nesse artigo, os autores analisaram 100.000 perfis do Friendster e do Myspace, deles inferiram uma fábrica do gosto e propuseram três aplicações: um mapa interativo de interesses (InterestMap), um sistema semântico (Ambient Semantic) para facilitar a interação entre duas pessoas estranhas a partir de seus gostos partilhados, e um espelho que reflete de forma dinâmica a identidade e o gosto dos indivíduos (IdentityMirror).


Interest Map; IdentityMirror; Ambient Semantics

Aplicações ainda mais sofisticadas e dinâmicas que os agentes de recomendação de livros e discos na amazom.com. Aplicações reveladoras sobre os cruzamentos entre redes sociais, composição de bancos de dados e perfis computacionais, cultura do consumo, fabricação das identidades pós-modernas e sistemas de monitoramento e vigilância no ciberespaço.
Segue resumo do artigo:
Unraveling the Taste Fabric of Social Networks
Hugo Liu, Pattie Maes, Glorianna Davenport
The Media Laboratory, Massachusetts Institute of Technology

Abstract
"Popular online social networks such as Friendster and MySpace do more than simply reveal the superficial structure of social connectedness—the rich meanings bottled within social network profiles themselves imply deeper patterns of culture and taste. If these latent semantic fabrics of taste could be harvested formally, the resultant resource would afford completely novel ways for representing and reasoning about web users and people in general. This paper narrates the theory and technique of such a feat—the natural language text of 100,000 social network profiles were captured, mapped into a diverse ontology of music, books, films, foods, etc., and machine learning was applied to infer a semantic fabric of taste. Taste fabrics bring us closer to improvisational manipulations of meaning, and afford us at least three semantic functions— the creation of semantically flexible user representations, cross-domain taste-based recommendation, and the computation of taste-similarity between people—whose use cases are demonstrated within the context of three applications—the InterestMap, Ambient Semantics, and IdentityMirror. Finally, we evaluate the quality of the taste fabrics, and distill from this research reusable methodologies and techniques of consequence to the semantic mining and Semantic Web communities."

Spock - indexando pessoas

Um dos principais focos da pesquisa de que trata esse blog é a constituição de uma nova maquinaria taxonômica através da elaboração de bancos de dados e perfis computacionais no ciberespaço. Argumentamos que diversos programas e sistemas de busca, produção e distribuição da informação na rede são também poderosos sistemas de coleta e classificação de informações sobre indivíduos, gostos, preferências, inclinações corportamentais etc. Em alguns artigos já indicados aqui, analisamos como mecanismos de busca (Google), redes sociais (Orkut) e sites de vendas (amazom.com) vêm construindo valiosos bancos de dados, alimentados pelas próprias ações e informações fornecidas pelos usuários destes serviços.

Em breve postarei um novo artigo sobre esse tema, levando em conta os novos dispositivos da Web 2.0. Antecipo, contudo, um dos serviços que merece atenção e destaque - o SPOCK. Seguindo a pista do Tim O'Reilly em entrevista mencionada no post anterior, fui conferir o serviço e ele é um excelente exemplo e uma atualização concreta das questões mencionadas acima. Trata-se de um mecanismo de busca pessoal, isso é, voltado para a busca e indexação de pessoas. Sim, o Google já permite buscar pessoas na rede, mas o Spock, por ser especializado e contar com ferramentas de classificação "colaborativas" oferece resultados bastante interessantes. Além de criar taggs para pessoas de modo automatizado, o Spock permite que os usuários adicionem suas próprias taggs, votem nas existentes, identifiquem relações entre pessoas e forneçam outros tipos de informação. Enfim, o sistema torna-se mais eficiente à medida que as pessoas o utilizam. Uma aplicação exemplar do modelo do que temos chamado de "vigilância colaborativa pu participativa", a qual é indissociável, como se pode ver, da constituição de bancos de dados, taxonomias e perfis também colaborativos.

Tim O'Reilly - web 2.0 e controle de dados

Em entrevista a WIRED, Tim O'Reilly fala do potencial da Web.20 (termo cunhado pelo autor) para coleta e controle de dados. A entrevista foi concedida na ocasião da Expo Web.20, realizada em abril de 2007 em São Francisco. Vejam a entrevista completa aqui e trecho abaixo:

Photo: Scott Beale/Laughing Squid

"WN: Are there any trends among the companies exhibiting at Web 2.0 Expo, the kinds of services and technologies being shown?

O'Reilly: Well, obviously this is a market with a lot of froth in it already. I have to say there are a lot of me-too products and companies. Yet another social network, of the 15th flavor -- that's common in every new technology revolution. There are imitators who have marginal improvements.

One of the companies that's going live on Monday is Spock, which is a people-search engine. It's really, really impressive. It's thinking about whether there are other classes of data to which search hasn't really been applied.

That goes back to a major theme of web 2.0 that people haven't yet tweaked to. It's really about data and who owns and controls, or gives the best access to, a class of data. Amazon is now the definitive source for data about whole sets of products -- fungible consumer products. EBay is the authoritative source for the secondary market of those products. Google is the authority for information about facts, but they're relatively undifferentiated.

Why did Google, for example, recently decide to offer free 411 service? I haven't talked to people at Google, but it's pretty clear to me why. It's because of speech recognition. It has nothing to do with 411 service, it has to do with getting a database of voices, so they don't have to license speech technology from Nuance or someone else. They want their own data stream.

WN: So you think that (control of data) is actually more characteristic of web 2.0 than social networking or Ajax (asynchronous JavaScript and XML) interfaces?

O'Reilly: Absolutely. Anybody who thinks that this is about Ajax is completely missing the boat.

I do think building rich internet applications is an important part of web 2.0. I don't want to dismiss it, because we are able to build richer application platforms today. But it's ultimately about network effects, and where do you build services that get better the more people use them? And it's also about the databases that get created as a result of those network effects.

As far as I'm concerned, web 2.0 is still in its really early stages, and the reason is because the data isn't all owned yet.

The network-effects play is about how you get increasing returns by everybody using your stuff, which is really what Microsoft did on the PC. Here we see it again, where these are winner-takes-all games. The internet looks like an open platform in the beginning, but once somebody gets a lead, their service gets better fast enough, if they've harnessed all the right levers, until it becomes a real barrier to entry.

Why, despite many attempts, have we seen nobody able to dethrone eBay? Well, it's because there are network effects at work in auctions. You have a critical mass of buyers and sellers. We're seeing that with Google AdWords -- it's just a bigger and better marketplace. There are these tipping points where these services really become monopolistic.

We're still trying to move people toward really understanding what that new world looks like. I don't think a lot of people are there. A lot of people still think, "Oh, it's about social networking. It's about blogging. It's about wikis." I think it's about the data that's created by those mechanisms, and the businesses that that data will make possible."