Data science increasingly uses machine learning techniques to develop predictive models for testing and analyzing volumes of data.
In this definition...
Why use data science?
Data science was born from a need to analyze unstructured data. Insights and trends collected can inform considerations and decisions.
Structured data filling rows and columns like Excel or Google Sheets was easy for entering, searching, comparing, extracting, and analyzing data in the past. The 1990s brought a move to predominantly unstructured or semi-structured data. Unstructured data includes business documents, emails, social media, customer feedback, webpages, open-ended survey responses, images, audio, and video.
Data scientist vs. data analyst
|“What is the data?”||“What does the data tell us?”|
|Focused on machine learning and algorithms||Focused on business administration|
|Developing operational models||Pre-processing and data gathering|
|In-depth programming knowledge||Scripting and statistical skills|
How is data science used?
Data science techniques reveal insights that inform decisions. Deciding to produce more of a specific product, building an office in a new location, altering the email marketing campaign with more CTAs (call-to-actions); if the data shows justification, the company can decide knowing where the data pointed.
Data science continues to evolve and hone its ability to test, manipulate, and utilize data from unstructured and semi-structured volumes.
Examples of data science
- Text analysis
- Mention mining
- Biometric analysis
Text analysis is the method of analyzing unstructured and semi-structured text for business insights. Be it five thousand customer surveys or three years of invoices, the application of data science to text analysis proves to outperform humans in less time and resources.
Mining for Mentions
On social media, mentions of organizations, brands, and products can inform digital marketing strategy. With data science applied, text analysis and machine learning can automate user insights on social media.
In an age of protecting human identities and the data associated with them, biometric analysis uses image, video, sensor, and biometric data to authorize users. From opening a smartphone to fingerprint identification and behavioral analysis, evaluating all of the unstructured data today would be impossible without data science.
History of data science
Data science came about as a term during the development of computers in the second half of the 20th century. Computer science was an upstart field of thought, while statistics had been a millennium in the making. For years, a debate over renaming fields of study ensued, but data science failed to catch fire until the 2000s.
Early timeline of data science as a term
|1962||Mathematician John Tukey proposes a field of study called data analysis.|
|1974||Computer scientist Peter Naur proposes data science replace the term computer science.|
|1985||Staticiation Chien-Fu Jeff Wu proposes data science replace the term statistics.|
|1990s||Knowledge discovery and data mining are used to describe data science.|
|1996||The International Federation of Classification Societies features data science as a topic.|
|1997||C.F. Jeff Wu proposes data science replace the term computer science (again).|
|1998||Hayashi Chikio proposes data science contain three aspects: data design, collection, and analysis.|