Data Mining and Cloud Computing

So, after the blog posts regarding "Smart Cities" and our discussion on Google, Facebook and their data mining efforts, I felt this was an appropriate supplementary piece. This article centers around the addition of sensors to current technologies to better predict what will happen; essentially, it performs a sort of behavior prediction. For example. General Electric (G.E.) is implementing LED street lamps that have sensors which can gather data. This data can work in many ways, such as determining if someone is moving under the light or not, to presumably turn it on or off. It can gather data on the heat emission and quality of the air in its immediate surroundings. With this technology in consumer households, people can gain insight on how their electric consumption.

This data can help the company save money of course and provide other benefits as well. One such benefit is to attract consumers to one's store by turning on/off a light with motion. It looks like this functionality can be extended to better understand and predict behavior. Consider how Google "mines" your link clicks and web searches to send targeted advertisements to your screen. Facebook uses a similar pattern, with it's "liked" pages. This is the crux of their business models.

Other companies are following suit with regards to the Facebook and Google model, such as Amazon. Amazon's makes $5 Billion (with a 'B') dollars per year by selling their computing services; this outsells the other services and products by Amazon. In addition to their computing services, they will begin to sell data analysis tools. Microsoft will be providing a cloud-based service and IBM will put Watson, it's Jeopardy! computer, in the cloud as well.

The idealist point-of-view sounds perfect: companies can access and analyze data from their sensors to better provide "services" to their customers. Amazon provides better shopping suggestions, Google and Facebook provide more relevant pages and ads. However, as we have seen in other blogs and readings, this is most definitely not always the case. In the Smart City blog, I already aired my concerns about companies having the potential to track your daily habiitys and through that, provide targeted advertisements, shopping suggestions, etc. via those sensors. It would have the potential to be extreme marketing: if a company has enough data and can analyze it for certain behaviors, it's literally the next best thing to having a someone follow you around and log your movements. Obviously companies don't have the time nor the resources to do this (plus they would be accused of breaking laws, so they snoop in a more "passive" way).

The thing is, the interconnection of digital and physical seems to be inevitable and facilitates the data collection and analysis efforts of tech companies. A Chinese computing manufacturer, Huawei, claims that by 2025 there will be 100 Billion (again, with a 'B') things connected to the cloud. Imagine: street lights, buses, cars, watches, refrigerators, traffic lamps, cash registers, etc. Everything connected and data being sent to and analyzed by big companies to better serve their business models.

http://bits.blogs.nytimes.com/2015/04/26/envisioning-a-future-when-sensors-are-scooping-up-data-on-everything/?ref=technology