Big data: Privacy issues should be resolved sooner or later

The report says big data raises serious privacy concerns to be resolved sooner or later.

The potential for big data is endless-and it can bring benefits and harm. A new White House report that describes the changing nature of big data delves into the data-related privacy and security topics.

Key concern: Large data is creating a large number of privacy topics that need to be dealt with in a timely manner, sooner or later.

The report's citation said: "One notable finding of this report is the large data analysis that uses personal information for housing, credit, employment, health, education and shopping malls, and has the potential to overshadow long-standing civil rights protections." The relationship between the Americans and the data, the opportunity and potential of her (his) will be expanded rather than diminished.

The report discusses a number of topics on privacy, including the following 5:

1. "To identify" is not always effective

Organizations often use privacy protection techniques to remove connection recognition from specific personal or device data. Unfortunately, the recognition technology is equally effective in connecting these data.

The report proposes: "A comprehensive range of data can lead to what some analysts call a" mosaic effect ", where personally identifiable information can be derived or inferred from a dataset that does not even include a personal identification code, focusing on depicting a person's image and what she or she likes.

As the technology for identifying anonymous data becomes increasingly powerful, how individuals manage their personal information and identities, or make decisions based on multiple data sets, is becoming confusing.

2. "Perfect personalization" can help to identify power

The report says that the integration of different types of http://www.aliyun.com/zixun/aggregation/13739.html "> unstructured data allows marketers to" accurately send the message that they want, before the consumer asks, The product or service sends the past "." Unfortunately, perfect personalization leaves room for subtle and less subtle recognition of pricing, services, and opportunities. ”

3. "Small" data create greater privacy threats

No matter how much talk about the potential infringement of individual rights over large data, most of the most common privacy risks today are related to "small data", such as hackers targeting financial fraud with personal bank information. "These risks are not always large, rapid or informative, and do not imply a complex analysis related to large data," the report said. The report argues that the protection of small data has been regulated by U.S. privacy laws, strong enforcement rules, and a global privacy-protection mechanism. Even if it is true, the last 2 articles show that there is much room for improvement in this area.

4. Predicting that medicine may lead to privacy disorder

"Predictive medicine" is a large, promising data application that delves into the patient's health and genetic information to predict whether she or she will have a particular disease and how receptive they are to specific treatments. The potential for data abuse here is enormous. For example, health information collected through "predictive medicine" may be used in medical decisions for people with similar genes, such as patients ' children.

"The current coverage and application of a healthy privacy framework are not well adapted to deal with these new developments or to promote research that drives new developments," the report said.

5. Conversely, privacy laws impede some important analysis

"Large data analysis allows data scientists to accumulate large amounts of data, including unstructured data, and identify anomalies and patterns," the report said. "One of the key privacy challenges in this discovery model is that you have to go to the sea to get a needle." In order to gain some insight, you need a certain amount of data. ”

So the problem comes out: Researchers can benefit from acquiring more sensitive genetic information datasets, but privacy laws limit her (him) access to the data. For example, a genetic researcher at the Broad Institute was unable to detect genetic mutations associated with 3,500 schizophrenia gene datasets, but obtained 35,000 cases of "statistically significant" results, the report said.