Data science also integrates domain knowledge from the underlying application domain (e.g., natural sciences, information technology, and medicine).[3] Data science is multifaceted and can be described as a science, a research paradigm, a research method, a discipline, a workflow, and a profession.[4]
Data science is often described as a multidisciplinary field because it draws on techniques from diverse areas, such as computer science, statistics, information science, and other subject-specific disciplines. Some researchers say that the combination of the different fields is similar to how information science was decades ago (Mayernik, 2023). These similarities help us understand how data science became its own field of study.[9]
Adata scientist is a professional who creates programming code and combines it with statistical knowledge to summarize data.[10]
Vasant Dhar writes that statistics emphasizes quantitative data and description. In contrast, data science deals with quantitative and qualitative data (e.g., from images, text, sensors, transactions, customer information, etc.) and emphasizes prediction and action.[13]Andrew Gelman ofColumbia University has described statistics as a non-essential part of data science.[14] Stanford professorDavid Donoho writes that data science is not distinguished from statistics by the size of datasets or use of computing and that many graduate programs misleadingly advertise their analytics and statistics training as the essence of a data-science program. He describes data science as an applied field growing out of traditional statistics.[15]
In 1962,John Tukey described a field he called "data analysis", which resembles modern data science.[15] In 1985, in a lecture given to the Chinese Academy of Sciences in Beijing,C. F. Jeff Wu used the term "data science" for the first time as an alternative name for statistics.[16] Later, attendees at a 1992 statistics symposium at theUniversity of Montpellier II acknowledged the emergence of a new discipline focused on data of various origins and forms, combining established concepts and principles of statistics and data analysis with computing.[17][18]
The term "data science" has been traced back to 1974, whenPeter Naur proposed it as an alternative name to computer science. In his 1974 book Concise Survey of Computer Methods, Peter Naur proposed using the term ‘data science’ rather than ‘computer science’ to reflect the growing emphasis on data-driven methods[19][6] In 1996, the International Federation of Classification Societies became the first conference to specifically feature data science as a topic.[6] However, the definition was still in flux. After the 1985 lecture at the Chinese Academy of Sciences in Beijing, in 1997C. F. Jeff Wu again suggested that statistics should be renamed data science. He reasoned that a new name would help statistics shed inaccurate stereotypes, such as being synonymous with accounting or limited to describing data.[20] In 1998, Hayashi Chikio argued for data science as a new, interdisciplinary concept, with three aspects: data design, collection, and analysis.[18]
In 2012, technologistsThomas H. Davenport andDJ Patil declared "Data Scientist: The Sexiest Job of the 21st Century",[21] a catchphrase that was picked up even by major-city newspapers like theNew York Times[22] and theBoston Globe.[23] A decade later, they reaffirmed it, stating that "the job is more in demand than ever with employers".[24]
The modern conception of data science as an independent discipline is sometimes attributed toWilliam S. Cleveland.[25] In 2014, theAmerican Statistical Association's Section on Statistical Learning and Data Mining changed its name to the Section on Statistical Learning and Data Science, reflecting the ascendant popularity of data science.[26]
Over the last few years, many colleges have begun to create more structured undergraduate programs in data science. According to a report by the National Academies, strong programs typically include training in statistics, computing, ethics, and communication, as well as hands-on work in a specific field (National Academies of Sciences, Engineering, and Medicine, 2018). As schools try to prepare students for jobs that use data, these practices become more common.[27]
The professional title of "data scientist" has been attributed toDJ Patil andJeff Hammerbacher in 2008.[28] Though it was used by theNational Science Board in their 2005 report "Long-Lived Digital Data Collections: Enabling Research and Education in the 21st Century", it referred broadly to any key role in managing a digitaldata collection.[29]
In data science,data analysis is the process of inspecting, cleaning, transforming, and modelling data to discover useful information, draw conclusions, and support decision-making.[30] It includesexploratory data analysis (EDA), which uses graphics and descriptive statistics to explore patterns and generate hypotheses,[31] andconfirmatory data analysis, which applies statistical inference to test hypotheses and quantify uncertainty.[32]
Typical activities comprise:
data collection and integration;
data cleaning and preparation (handling missing values, outliers, encoding, normalisation);
Recent studies indicate that AI is moving towards data-centric approaches, focusing on the quality of datasets rather than just improving AI models. This trend focuses on improving system performance by cleaning, refining, and labeling data (Bhatt et al., 2024). As AI systems grow larger, the data-centric view has become increasingly important.[37]
A cloud-based architecture for enabling big data analytics. Data flows from various sources, such aspersonal computers,laptops, andsmart phones, through cloud services for processing and analysis, finally leading to variousbig data applications.
Cloud computing can offer access to large amounts of computational power andstorage.[38] Inbig data, where volumes of information are continually generated and processed, these platforms can be used to handle complex and resource-intensive analytical tasks.[39]
Some distributed computing frameworks are designed to handle big data workloads. These frameworks can enable data scientists to process and analyze large datasets in parallel, which can reduce processing times.[40]
Data science involves collecting, processing, and analyzing data which often includes personal and sensitive information. Ethical concerns include potential privacy violations, bias perpetuation, and negative societal impacts.[41][42]
Ethics education in data science has grown to encompass both technical principles and more expansive philosophical questions. Research indicates that data science ethics courses are increasingly integrating human-centric topics, including fairness, accountability, and responsible decision-making, thereby connecting them to enduring discussions in moral and political philosophy (Colando & Hardin, 2024). The goal of this method is to help students understand how data-driven technologies affect society.[43]
Machine learning models can amplify existing biases present in training data, leading to discriminatory or unfair outcomes.[44][45]Another area of data science that is growing is the push for better ways to cite data. Citing datasets makes it easier for other researchers to understand what data was used and for studies to be repeated (Lafia et al., 2023). These practices give the people who collect and manage data the credit they deserve, which is becoming more important in modern research.[46]
^abJames, Gareth; Witten, Daniela; Hastie, Trevor; Tibshirani, Robert (2017).An Introduction to Statistical Learning: with Applications in R. Springer.ISBN978-1-4614-7137-0.
^O'Neil, Cathy; Schutt, Rachel (2013).Doing Data Science. O'Reilly Media.ISBN978-1-4493-5865-5.
^CRISP-DM 1.0: Step-by-step data mining guide (Report). SPSS. 2000.
^Armbrust, Michael; Xin, Reynold S.; Lian, Cheng; Huai, Yin; Liu, Davies; Bradley, Joseph K.; Meng, Xiangrui; Kaftan, Tomer; Franklin, Michael J.; Ghodsi, Ali; Zaharia, Matei (27 May 2015)."Spark SQL: Relational Data Processing in Spark".Proceedings of the 2015 ACM SIGMOD International Conference on Management of Data. ACM. pp. 1383–1394.doi:10.1145/2723372.2742797.ISBN978-1-4503-2758-9.
.jpg)
