問27
頻出度★★★
蓄積された販売データなどから, 天候と売れ筋商品の関連性などの規則性を見つけ出す手法を表す用語はどれか。
平成28年度 秋期 ITパスポート試験 問27
ストラテジ系企業活動 › 業務分析・データ利活用
This question's explanation and your answers so far.
解説
問 27
Incorrectア
A data warehouse is a “data repository” that organizes and integrates data collected from various business systems in time series and stores it for analysis. It is a mechanism for accumulating the data that forms the basis of analysis, not a technique for discovering regularities from it, so it is wrong.
Incorrectイ
Data processing refers to general data handling that routinely processes input data—aggregating, sorting, updating, and so on. It is not an analytical technique for discovering hidden regularities or correlations, so it is wrong.
- ウ
正解 The correct answer is ウ, “Data mining.” Data mining is a technique that uses methods such as statistical analysis and machine learning to dig out regularities, correlations, and patterns that people find hard to notice from large volumes of accumulated data (mining = digging out). Finding hidden relationships between seemingly unrelated items from sales data—such as the relationship between weather and best-selling products in the question—is exactly the role data mining plays, and it is widely used in marketing and demand forecasting.
Incorrectエ
Data modeling is the work of designing and expressing the structure of the data handled in business operations and the relationships among items using ER diagrams and the like; it refers to a database design step. It is not a technique for mining regularities from already-accumulated data, so it is wrong.
ここがポイント
- Data mining = a technique that uses statistical analysis, machine learning, and the like to discover hidden regularities, correlations, and patterns from large volumes of accumulated data (e.g., co-purchase patterns found by basket analysis, correlation between weather and sales).
- Division of roles among easily confused similar terms: data warehouse = a repository for storing data for analysis, data mining = a technique for discovering knowledge from it, data modeling = designing the data structure.
用語