本文へ移動
×
確認テスト
外国語・Explainer: algorithms and recommendations(5問)
次の英文資料の明示情報と一致する英文を選べ。 A recommendation system does not simply “know” what a user likes. It predicts interest from signals such as viewing time, clicks, and similarities among users. Those signals are incomplete: a long viewing time may indicate confusion rather than enjoyment. Moreover, recommendations shape the behavior later used as data. If a platform repeatedly highlights one topic, users may click it more often simply because it is visible. Designers therefore need to test alternative explanations and give users meaningful control.
ア
A click always means that content was hidden.
イ
A long viewing time can reflect confusion rather than enjoyment.
ウ
Similar users never provide useful signals.
エ
Platforms cannot record viewing time.
戻る
1 / 5
次へ