Data, algorithms and digital power · 数据、算法与数字权力
| English | 中文 | Pinyin · 拼音 |
|---|---|---|
| algorithm/ˈælɡərɪθəm/ | 算法 | suàn fǎ |
| digital footprint/ˈdɪdʒɪtl ˈfʊtprɪnt/ | 数字足迹 | shù zì zú jì |
A decision you can investigate
- A school app recommends study groups using attendance and quiz records.
- A useful prediction may still expose information or reproduce unequal opportunities.
一个你可以调查的决策
- 一款学校应用程序根据出勤率和测验记录推荐学习小组。
- 一个有用的预测仍可能泄露信息或再生产不平等机会。
Build the explanation
- Data are recorded values; information gives them meaning in a context. An algorithm 算法 specifies a process for producing outputs.
- Ask who supplies data, who chooses the objective, who can challenge a decision and who is excluded.
构建解释
- 数据是记录的值；信息赋予其在特定背景下的意义。算法 algorithm 指定了生成输出的一系列过程。
- 询问谁提供数据，谁设定目标，谁能质疑决策以及谁被排除在外。
Match the terms to their precise meanings. · 将术语与其精确含义匹配。
Use these definitions in the particular context of Data, algorithms and digital power. · 在数据、算法和数字权力的特定语境中使用这些定义。
Work through the evidence
- A missing quiz can mean absence, connection failure or refusal to participate. Treating every missing value as low ability changes recommendations.
- Investigate missing-data rules and compare outcomes across users before describing the system as fair.
梳理证据
- 缺失的测验可能意味着缺勤、连接失败或拒绝参与。将所有缺失值视为低能力会改变推荐结果。
- 在描述系统公平之前，调查缺失数据的规则并比较不同用户的结果。
Why is replacing every missing quiz with zero risky? · 为什么将所有缺失测验替换为零是危险的？
A missing quiz can mean absence, connection failure or refusal to participate. Treating every missing value as low ability changes recommendations. Investigate missing-data rules and compare outcomes across users before describing the system as fair. · 缺失的测验可能意味着缺勤、连接失败或拒绝参与。将所有缺失值视为低能力会改变推荐结果。在描述系统公平之前，调查缺失数据的规则并比较不同用户的结果。
Test the limits
- A prediction does not establish a cause. High average accuracy can conceal unequal errors across groups.
- Explain the specific harm and stakeholder rather than merely writing privacy is important.
测试极限
- 预测不能确立因果关系。高平均准确率可能掩盖不同群体间的不平等误差。
- 说明具体的伤害对象和相关利益方，而不仅仅是写“隐私很重要”。
Which caution belongs to this particular task? · 哪项注意事项适用于此特定任务？
A prediction does not establish a cause. High average accuracy can conceal unequal errors across groups. Explain the specific harm and stakeholder rather than merely writing privacy is important. · 预测不能确立因果关系。高平均准确率可能掩盖不同群体间的不平等误差。说明具体的伤害对象和相关利益方，而不仅仅是写“隐私很重要”。
The explanation in this lesson makes a conditional claim; relevant context and evidence still matter. · 本课的解释提出了条件性主张；相关背景和证据仍然重要。
Explain the specific harm and stakeholder rather than merely writing privacy is important. · 解释具体的伤害和相关利益方，而不仅仅是写出“隐私很重要”。
Apply and explain your answer
- Why is replacing every missing quiz with zero risky?
- It treats different causes of missing data as evidence of low ability.
应用并解释你的答案
- 为什么用零替换所有缺失的测验存在风险？
- 它将缺失数据的不同原因都作为低能力的证据。
Choose the two statements supported by this lesson. · 选择本课支持的两项陈述。
The concept and worked evidence support these claims; the stated limits rule out the universal shortcut. · 概念和推演证据支持这些主张；明确的限制排除了普遍捷径。
Use the terms precisely
- algorithm: A specified process for transforming inputs into outputs.
- digital footprint 数字足迹: Traces left by a person's digital activity.
精确使用术语
- 算法 algorithm：指定的一组将输入转换为输出的过程。
- 数字足迹 digital footprint: 个人数字活动留下的痕迹。
A missing quiz can mean absence, connection failure or refusal to participate. Treating every missing value as low ability changes recommendations. Investigate missing-data rules and compare outcomes across users before describing the system as fair.
A prediction does not establish a cause. High average accuracy can conceal unequal errors across groups. Explain the specific harm and stakeholder rather than merely writing privacy is important.
Data are recorded values; information gives them meaning in a context. An algorithm specifies a process for producing outputs.
缺失的测验可能意味着缺勤、连接失败或拒绝参与。将所有缺失值视为低能力会改变推荐结果。在描述系统公平之前，调查缺失数据的规则并比较不同用户的结果。
预测不能确立因果关系。高平均准确率可能掩盖不同群体间的不平等误差。说明具体的伤害对象和相关利益方，而不仅仅是写“隐私很重要”。
数据是记录的值；信息在特定背景下赋予这些值意义。算法规定了生成输出的过程。