Everyone generates information as a by-product of their activities. Wear patterns indicate favorite couches, and fingerprints reveal objects that have been touched. Data exhaust is unstructured information that is a by-product of digital activity. Unlike worn couches that get thrown out or fingerprints that are cleaned off, recorded digital activity is stored for very long periods of time. This unstructured subset of big data can reveal valuable insights about people, and the nature of these insights depend on the questions asked. Search Engine Use One company that has hugely benefited from the left over digital exhaust of online activity is Google. Through its access to enormous amounts of information, it is continually evolving and improving its services. One challenge that Google faces is the manipulation of its search results. At first it relied primarily on the linking structure of the Internet to identify worthwhile websites. While this negated the manipulative practice of keyword stuffing, it fared less well when people learned how to point links at their own websites and thus manipulate Google’s ranking algorithm. Google has taken steps to correct this problem by making use of the digital exhaust of traffic behavior when it interacts with websites. Data such as length of stay and other indicators of interaction now supplement incoming links when Google assesses the quality of a site. Medical Use One doesn’t have to be a search engine giant to benefit from data exhaust. Ginger IO, a Boston based company, performs behavioral analytics on the data exhaust of cell phone users who have chronic medical conditions. This information allows doctors to remotely monitor the condition of these people who are their patients. One of Ginger IO’s apps can predict the onset of depression two days in advance. Depression causes people to stop taking care of themselves such as taking their medications. For some types of patients such as diabetics, not taking medication can have serious and even fatal consequences. How can a cell phone app predict depression? The app can monitor the frequency of cell phone use as well as a person’s physical movements through its GPS tracking. When a person gets depressed, they tend to withdraw from their normal activities and interactions. The app senses this through decreased cell phone activity and physical movement. Most people are creatures of habit and go to the same places and also have predictable cell phone habits. The app is familiar with this baseline behavior and reacts to the onset of depression by calling the person’s doctor and social network. Other Uses Mining data exhaust has countless possible applications. Employment changes can be inferred from changes in purchasing patterns. Data from mobile payment systems can be used to infer credit risk. Malaria outbreaks in third world countries can be predicted from the movements of large numbers of people through their cell phones. The relative affluence of people can be deduced by their spending patterns. Dynamic pricing can be based on a person’s online activity as well as cell phone usage. The possibilities of mining the data exhaust left behind by digital activity seem to be enormous. However, the possibility of controversy may also exist for some types of data mining. For more information about big data and its storage, please contact us.