Professor Thomas Davenport of Babson College, Harvard Business School, and the MIT Sloan School of Management gave the keynote address titled Four Eras of Analytics and Data Science at the Open Data Science Conference East 2017 in Boston.
Prof. Davenport’s speech explored the history of data analytics from a commercial standpoint, from the 1970s to the present, and divided the activity into four eras.

In comparison to the extensive arcs of biology, chemistry, and other sciences, data science has a relatively short history. Nonetheless, this history is rich in its own right, coming from a collection of trailblazers who stand in stark contrast to the academicians who founded the study of the physical world decades before. Continue reading to learn how we progressed from “back room” analysts to the “sexiest job of the twenty-first century.”

Artisanal Analytics is the first era.

Prof. Davenport believes that data analytics began in earnest in 1975 with what he refers to as “artisanal analytics.” Using small-scale, structured datasets, this methodology was primarily designed to produce insights for internal decision-making.

The analyst in the 1970s served a fundamentally different position in the company than the modern data scientist. Analysts were not producing reusable customer-facing tools. Rather, they focused on prediction models based on human hypotheses, which took a long time to perfect. This meant that, in Dr. Davenport’s words, the analyst was more of a “back room” support member than a front-line member.

Information Transformation Service provides  web scraping Services that provides high-quality structured data to improve business outcomes and enable intelligent decision making,their Web scraping service allows you to scrape data from any websites and transfer web pages into an easy-to-use format such as Excel, CSV, JSON and many others.

This type of analysis isn’t dead, according to Davenport, but it’s been pushed aside by other methodologies capable of uncovering insights without human participation, such as machine learning. Nonetheless, artisanal analysts made substantial contributions that are being utilised today, including as business-oriented statistics and basic visualisation.

Big Data Analytics in the Second Era

As Silicon Valley grew in popularity in the late 1990s and early 2000s, so did the amount and variety of data available to analysts. As a result of the new issues that this shifting landscape presented, a new title emerged: data scientist.

Data input is a briskly growing sector in Pakistan. Information Transformation Services can meet a variety of data entry services competently and professionally. ITS will not only provide various data entry services to customers, but will also offer some creative and customized solutions to meet your needs

“I must admit that when I first heard about big data and data scientists and so on, I wasn’t completely sure that this was anything that different from the sort of analytics that I had been talking about and writing about,” Davenport said. As a result, I began to research them.”

Dhanurjay “DJ” Patil, who would go on to become the first Chief Data Scientist of the US Office of Science and Technology Policy, began working with Davenport. Patil’s comments on the distinction between analyst and data scientist proved enlightening in their discussions. Patil would remark, according to Davenport, that “data scientists need to be on the bridge… right up there with Captain Kirk,” taking responsibility of the decision-making process rather than supporting leaders from behind.

“Supporting management decision-making?” says Davenport. That’s where the dead zone is.”

Data Economy Analytics in the Third Era

Another significant change occurred prior to and around the year 2013. As large IT companies discovered new ways to manage their huge databases, they also discovered new ways to commoditize them. They began to sell the data they were collecting from consumers, in addition to constructing products around the databases they kept – the strategy firms had used for nearly forty years, according to Davenport.

“Industrialized decision-making at scale,” as Davenport put it, became the new means of using data, reacting to changes as quickly as new data arrived. Next to Davenport’s first period, where models were unresponsive to new information — especially since so little new information was entering the picture at any given time – business analytics became practically unrecognisable. Companies could now not only collect information from users and sell it as a commodity, but they could also adjust their strategy on the fly, without human intervention.

Autonomous Analytics is the fourth era.

According to Davenport, in the last year, we’ve entered a new era of analytics marked by a greater role for autonomous decision-making — certainly the loosest definition of artificial intelligence.

Machines, under this concept, not only execute the analysis, but also act on the findings, making judgments faster and more efficiently than humans.

Davenport’s research differs from that of many automation alarmists who argue that AI and machine learning will render existing employment obsolete. Data scientists who refuse to update their toolkits, in his opinion, are in the crosshairs. “The only people who will lose their jobs are those who do not accept these new technologies,” he says.

Prof. Davenport underlined that the modern data scientist is responsible for mastering the complete range of techniques, from those used by ‘back room’ analysts in the 1970s to the sophisticated deep learning approaches currently in use. Fortunately, many of these approaches logically build on one another, making learning new techniques quite simple if you are already familiar with the precursors.

Prof. Davenport’s entire video lecture may be seen here: Four Eras of Analytics and Data Science.