Data Mining

The implications of our time browsing the Internet

Source: Flickr Creative Commons

Computer programs can extract an extraordinary and frightening amount of data from our web activity. Following the digital revolution and the proliferation of digital record keeping, the information generated by our online activity has potentially great benefit, but leaves serious questions unanswered regarding our confidence in digital record keeping.

As the world becomes increasingly interconnected and technological adoption rises in emerging markets, data from online activity poses great promise and peril. According to studies from the Pew Research Center, there has been a significant increase in the number of internet and smartphone users in developing nations. In 2013, across 21 emerging and developing countries, a median of 45 percent of respondents used the internet or owned a smartphone. In 2015, the median rose to 54 percent, with much of the increase driven by emerging economies like Malaysia, Brazil, and China. Through machine learning algorithms, we can now make well calculated and informed policies with previously unimaginable amount of data from observing the activity of the newly connected. With the massive increase in smartphone ownership among residents of developing nations, passively collected digital data could come to replace household surveys in assessing social and economic well-being. Information from online data collection could thus be translated into more effective policies and better service delivery.

Organizations such as Google, Amazon and Apple already collect data and invest heavily in technologies like big data, supercomputers and artificial intelligence and can tailor information directly to users. Facebook, for example, measures users’ links clicked, and by studying billions of data points it can determine the type of news a user might “like” and push that to their Facebook feed. Though this is an effective way to personalize the web, society must consider the ethics and implications of tailoring information, as the loss of reliable and authoritative information sources can inflame emotions and undermine reason.

In China, the nation is experimenting with “citizen scores”, which will rank citizens and could determine the conditions that citizens get loans, jobs, or travel visas. People’s internet activity and the behavior of their social contacts could potentially be monitored. Based on different data points, a user could be categorized over petabytes of information and millions of other users. Conceivably, without needing any consent or active inputs by the user, computer programs could guess preferences, estimate income, profession, educational level, and numerous other details. Algorithms can learn to recognize life events like pregnancies or even whether you’re thinking of getting married. GPS location can determine exact locations, home addresses, and commute time. Programs even have the capacity to predict your age based on mouse movement; younger people have more precise movements than older people.

We begin to better understand complex relationships in the world by applying algorithms and statistical methods to interpret the figures, names, and other quantities of online activity data. However, all this necessitates democratic technologies for greater transparency, trust and systems that are compatible with democratic principles. This requires decentralized information systems, commitments to open data strategies, reduced distortion of information and even granting citizens the right to get a copy of personal data collected on them.

Utilizing information aggregated through online activity and translating it into big data, we can quickly and intelligibly shift through millions of data points to extract meaningful relationships in complex economies, national systems, and human behaviors. Data gathering through online activity and employing big data techniques can be an extremely effective method for understanding and describing the world, yet as citizens, we must determine the extent and limit to the data we allow to be collected on us, how it’s collected, and who could have access to it. Policies and regulations for which governments and private industry must abide should be ahead of this technological change, as these algorithms are already creeping into every aspect of our lives.