第1章コンピューターサイエンス入門

By the end of this section, you will be able to:

• Differentiate between discovery and invention

• Describe how science, mathematics, and engineering each play a role in computer science

• Discuss how data science, computational science, and information science each relate to computer science

• Explain why the various areas of computer science are synergistic

Computer science is an incredibly diverse field not because of what it can achieve on its own but because of how it contributes to every other field of human knowledge and expertise. From its early days, it was understood that there would be cross-collaboration between computer scientists and colleagues in other disciplines. Today, almost all modern technologies either depend on computer technologies or benefit significantly from them. Computer technologies and the study of computer science have reshaped almost all facets of life today for everyone.

Across business, financial, governmental, scientific, and nonprofit workplaces, millions of people are programming, and most of the time, they don’t even know it! A spreadsheet is an example of a data-centric programming environment where data is organized into cells in a table. Instead of presenting programs as primarily about algorithms, spreadsheets present programs as primarily about data. Spreadsheets are often used for data analysis by offering a way to organize, share, and communicate ideas about data. Spreadsheets are uniquely effective and accessible because they allow for the visual organization of data in whichever structure makes the most sense to the user. Instead of hiding data behind code, spreadsheets make data as transparent and up to date as possible.

Although spreadsheets make computation accessible for millions of people across the world, they have several shortcomings. Unless limits are removed, many popular spreadsheet software products such as Microsoft Excel may have a limitation to the number of rows of data they can store that is less than the data of modern computers. One such example occurred in October 2020 when Public Health England failed to report 15,841 positive cases of COVID-19 in the United Kingdom due to mismanaged row limits in the spreadsheet used. This shortcoming attests not only to the technical limit on the number of rows supported by spreadsheets, but also to the design limitations of software that fails to communicate data loss, irregularities, or errors to users. Errors in spreadsheet software data entry can often go unnoticed because spreadsheets do not enforce data types. Cells can contain any content: numbers, currencies, names, percentages, labels, and legends. The meaning of a cell is determined largely by the user rather than the software. Spreadsheets are an expressive and accessible technology for data analysis, but this creative power that spreadsheets afford to users is the very same power that limits spreadsheets as a data management and large-scale data analysis tool. The more data and the more people involved in a spreadsheet, the greater the potential for spreadsheet problems.

The interdisciplinary field that applies computing to managing data and extracting information from data is called data science. Data scientists are practitioners who combine computing and data analysis skills with the domain knowledge specific to their field or business. The demand for data scientists is becoming increasingly important as more and more research and business contexts involve analyzing big data, or very large datasets that are not easily processed using spreadsheets. These datasets often involve high-volume measurements of user interactions on the Internet at a very fine grain, such as tracking a customer’s web browser history across an online storefront. Data scientists can then analyze browser patterns using machine learning methods in order to recommend related products, target advertisements to specific customers over social media, and reengage customers over email or other means. Machine learning (ML) is a subset of artificial intelligence that relies on algorithms and data to make it possible for artificial intelligence to learn, actually mimicking the way humans learn. For example, ML is used to identify fraudulent versus legitimate banking transactions. Once a computer learns how to distinguish fraudulent transactions, it can be alert and call attention to suspicious banking activity.