IPA IP 2026 Question 16
A question IPA Japan published in 令和8年度 ITパスポート試験 公開問題 問16, translated into English. ITPEC is based on the same exams.
In supervised learning, which of the following is the most appropriate term that represents the task of attaching the information that serves as the correct answer?
Explanation
Annotation — attaching the correct answers (labels) to the data that a supervised learning model learns from.
Supervised learning trains a model on pairs of input and correct output. Before training, people (or tools) tag each item with its answer: an image with “cat” or “dog”, a sentence with its sentiment, the position of each object in a photo. This tagging is annotation, and the tagged data become the training data. The quality of the annotation sets a limit on the accuracy of the model.
Answer (a)
Why not others:
- •(b) — encoding converts data into another representation (character codes, compressed video, categories into numbers); it adds no correct answers
- •(c) — data cleansing corrects or removes errors, duplicates and inconsistencies in data so that it can be used
- •(d) — filtering picks out or removes data that meet a condition
Key rule: Supervised learning needs labelled data, and labelling it is annotation; unsupervised learning finds patterns in data with no labels.
Basic Theory · AI-written, may contain errors
This question comes from the past questions IPA Japan publishes, translated into English by ITPEC Practice. ITPEC Practice is an independent study tool and is not affiliated with IPA or ITPEC. See the original paper (PDF) or Report an issue.