Many fields today are facing the Danger of Big Data, but the real question is- Is Big Data Dangerous and how it affects us. Generally these large volumes of data that make possible the analysis of very large sets of heterogeneous data. This data can be processed and exploited for the purpose of deriving intelligible and relevant information.
It is certain that the treatment of these masses of data will play a key role in the society of tomorrow. Nevertheless, while this is a big step forward, is this ever-growing collection of data not a risk to individuals or a danger to society?
Big Data Definition
Big Data is characterized by the "3V" rule, Volume, Variety, and Speed. The volume corresponding to the large volume of raw data and the variety to the heterogeneous data that make up Big Data. Speed is the speed at which Big Data is generated, sometimes continuously, which means processing it quickly, even in real time.
Today Big Data is rooted in fields as diverse as science, marketing, customer services, sustainability, transportation, health, and education. This is to allow companies or the state to better understand their environment and their users. In the case of companies, for example, the danger of Big Data offers many possibilities that make it possible to better analyze the needs of their customers, to better know them and why not even to detect trends. In the administrative framework, Big Data allows the state to better detect fraud but also to understand the concerns of citizens. But the question often hunts authority is that Big Data Dangerous for citizens and data collected by them is safe.
Big Data Dangerous
This burgeoning technology offers a tremendous number of benefits and is a breakthrough and a promise for business. But, like any technological advance, Big Data has its limitations and this leads us to question the potential major risks is that can Big Data Dangerous for us.
First of all, it is an ethical risk, let us recall the "Prism" scandal, the American program of electronic surveillance by collecting information from the Internet. The transmission of data from this surveillance program, which has been discovered and which the American administrative service is minimizing, totally calls into question the respect of the private life of individuals.
The predictive model, which can be used by insurance companies to refuse treatments or customers, also comes to be placed in the list of risks of Big Data. Moreover, the League of Human Rights has opened a seminar on the exploitation of data and algorithms in the field of insurance that could lead to excesses such as significant risks of discrimination. In a completely different field, one might even wonder if dangerous big data and health are compatible since the exploitation of personal data would ultimately go against medical confidentiality.
Exploitation of Big Data
Finally, there is the problem of data security since it is very difficult to store such a mass of data without any risk of intrusion. As for the freedom of individuals, does Big Data not hinder our freedoms? The borderline between individual freedoms and the exploitation of big data is quite narrow, and it is tricky to take advantage of these data while respecting the freedoms of each. there is the problem of data security since it is very difficult to store such a mass of data without any risk of intrusion.
As for the freedom of individuals, does Big Data not hinder our freedoms? The borderline between individual freedoms and the exploitation of big data is quite narrow, and it is tricky to take advantage of these data while respecting the freedoms of each. there is the problem of data security since it is very difficult to store such a mass of data without any risk of intrusion.
The danger of Big Data
As for the freedom of individuals, does Big Data not hinder our freedoms? The borderline between individual freedoms and the exploitation of data is quite narrow, and it is tricky to take advantage of these data while respecting the freedoms of each.
Facebook relaunches the debate with the investigation that targets, suspecting the social network to have delivered data to a British company for political purposes. The case of the Cambridge Analytica, a private Anglo-American company, would position itself as a company capable of influencing voters through a mix between Big Data, psychological profiling, and microtargeting.
In this case, there is an obvious rift in the danger of big data, either from Facebook or from this company. This confirms once again the problem posed by the security of personal data. Indeed, there is in one case or another, if this investigation turns out to be proven, an abuse of the exploitation of personal data.
Technology to watch closely
Nevertheless, despite potential loopholes, Big Data marks a revolution in the digital world, and increasingly intelligent algorithms are fully representative of Big Data. Facebook, for example, has developed a feature that allows Messenger to look in our album the photos we took with our Facebook friends and invites us to share them with them. This facial recognition algorithm is a huge breakthrough on a computer scale. With the advent of algorithms, data collections, this type of analysis is now possible.
This is even more successful than before the photos and videos were considered "unstructured" data. Thus, the computer did not have the necessary references to determine the identity of the person or the place where the photo was taken. Today the photos can be analyzed by algorithms that gives them a real structure. A whole new era of data is emerging and social networks are the first to be concerned and lead us to dangerous big data.
Big Data has had some success at Target and Amazon in particular, who have used Big Data in order to gain competitive advantages in sales. In his book, Nate Silver mentions how the massive use of data in computing has improved the quality of weather forecasts, especially over the past 25 years.
However, Silver also points out that weather forecasting is limited in use and accuracy when applied to dates farther than one week apart. It's a way for Silver to show the power and the limits of Big Data. Certainly, there are some successes of Big Data, but you will notice that the same examples as those just presented are common.
Finally, we can not deny that Big Data can be dangerous for individuals and society because of the excesses and flaws it can induce. On the other hand, Big Data marks an undeniable advance and it is necessary to reinforce the monitoring of the exploitation of these big data to avoid drifts.
Dangers of Big Data: is Big Data doing more harm than good?
Big Data plays a major role in many companies today. Data analysis algorithms are frequently used to assist decision-making, and their advice is rarely called into question. Yet the dangers of Big Data are real. Analytical technologies are far from foolproof. Find out why, if misused, Big Data can do more harm than good.
Dangers of Big Data: Ghost Data
The Big Data process is always the same. The data is extracted from huge databases, then analyzed by complex algorithms before finally being used for decision making. In the eyes of business leaders, the information they reveal is taken for gospel words. However, the source of these data is not always very reliable. In many firms, data is collected by low-skilled employees and may be wrong from the outset. It is impossible to eradicate all errors. However, these errors can pose concrete problems within the company. For example, ghost stocks can lead a firm to make bad purchasing or marketing decisions. A recent in the field of retail has proven that 65% of inventories are imprecise.
Big Data Dangers: Big Data and Layoff
In many areas, Big Data is commonly used. For example, analytics technologies are frequently used to measure employee performance across many organizations. One example is the story of Sarah Wysocki, a schoolteacher adored by her students, parents, and colleagues, who was fired because her performance was deemed wrong by an algorithm. The judgments carried by Big Data algorithms are generally without appeal, even though the methods used by these algorithms are very rarely revealed.
Big Data Dangers: Decisions Based on the Past
To recruit the best employees, many companies rely on Big Data. They develop a predictive model based on the performance of previous people recruited in the past, for example, to determine which schools train the best professionals. This method may seem relevant, but it is far from foolproof. Employee samples are often too weak to draw relevant statistical inferences, and it is sufficient if a student is better or worse than the average to distort the data. Thus, predictive Big Data models often have absolute confidence in the past and lead to erroneous predictions.
Dangers of Big Data: algorithms are not sufficiently tested
Unlike Wall Street elite mathematicians, big data algorithm developers do not lose money when their analytic models are wrong. It is the companies that pay the fees. At a time when data is increasingly influencing companies, it is necessary to better test algorithms, to consider which elements are taken into account, and which ones are not. A stricter look must be focused on machine intelligence. It is essential to know how data is collected, how conclusions are drawn, and whether they actually improve business activity. If not, Big Data can do more harm than good.
Also published on Medium.