📚 Interdisciplinary Integrated Question Practice for CCEA A-Level Computer Science | CCEA A-Level计算机跨学科综合题型训练
Cross-disciplinary questions in CCEA A-Level Computer Science require you to apply computational thinking, programming, data structures and systems analysis to problems drawn from mathematics, science, business and design. This article provides targeted practice scenarios with paired English and Chinese explanations, strengthening both your subject knowledge and your ability to tackle exam-style integrated questions.
CCEA A-Level计算机科学中的跨学科题目要求你将计算思维、编程、数据结构和系统分析应用到来自数学、科学、商业和设计等领域的问题中。本文提供针对性的练习场景，并配有中英文对照解释，既能巩固你的学科知识，也能提升你应对考试中综合题型的能力。
1. Mathematics – Recursive Sequences and Big O Analysis | 数学——递归数列与Big O分析
A sequence is defined by T(1)=1, T(n)=T(n-1)+3n² for n>1. A student implements a recursive function without memoisation. Analyse the time complexity of this implementation and suggest how an iterative approach can reduce it to O(n).
某数列定义为 T(1)=1, T(n)=T(n-1)+3n² (n>1)。一位学生用无记忆化的递归实现了该函数。请分析该实现的时间复杂度，并说明迭代方法如何将其降至 O(n)。
The recursive version computes the same sub‑problems repeatedly, leading to O(2ⁿ) worst‑case behaviour if implemented naively as tree recursion. By contrast, a single loop that accumulates the sum of 3n² from 1 to n executes in O(n) time. The closed‑form formula n(n+1)(2n+1)/2 further reduces it to O(1).
递归版本会重复计算相同的子问题，如果简单地实现为树形递归，最坏情况会达到 O(2ⁿ)。相比之下，用一个循环从1到n累加 3n² 只需 O(n) 时间。而闭式公式 n(n+1)(2n+1)/2 更可将其降至 O(1)。
2. Physics – Numerical Integration in Projectile Simulation | 物理——抛体模拟中的数值积分
A games engine must update the position of a projectile every 16 ms using Forward Euler. Given acceleration due to gravity g, initial velocity (vx, vy) and position (x, y), outline the update equations and discuss how reducing Δt affects accuracy and floating‑point errors.
某游戏引擎须每隔16毫秒用前向欧拉法更新抛体的位置。已知重力加速度 g、初速度 (vx, vy) 和位置 (x, y)，写出更新方程，并讨论减小 Δt 对精度和浮点误差的影响。
The equations are: vx ← vx, vy ← vy + g·Δt, x ← x + vx·Δt, y ← y + vy·Δt. Halving Δt halves the local truncation error but doubles the number of steps, which may accumulate total rounding error. Choosing an appropriate step size requires balancing truncation error and machine epsilon.
更新方程为：vx ← vx, vy ← vy + g·Δt, x ← x + vx·Δt, y ← y + vy·Δt。将 Δt 减半可使局部截断误差减半，但步数翻倍，从而可能累积更多的舍入误差。选择合适的步长需要在截断误差与机器精度之间取得平衡。
3. Biology – DNA Pattern Matching with String Algorithms | 生物——DNA模式匹配与字符串算法
A bioinformatics tool searches a DNA sequence of length 10⁶ for a short motif like ‘ACGTAC’. Compare the efficiency of a naïve O(n·m) search with the Knuth–Morris–Pratt algorithm, and describe how the failure function avoids backtracking.
某生物信息学工具在长度为10⁶的DNA序列中搜索诸如’ACGTAC’的短基序。比较朴素 O(n·m) 搜索与KMP算法的效率，并描述失效函数如何避免回溯。
Naïve search shifts the pattern by one position after a mismatch, re‑examining characters already compared. KMP precomputes a failure array π where π[i] gives the length of the longest proper prefix that is also a suffix for the pattern[0..i]. On a mismatch at position j, the pattern shifts by j – π[j-1], preserving O(n+m) worst‑case time.
朴素搜索在发生失配后将模式移动一个位置，重新检查已经比对过的字符。KMP预先计算失效数组 π，其中 π[i] 表示模式[0..i]中最长的同时也是后缀的真前缀长度。在位置 j 失配时，模式跳过 j – π[j-1] 个位置，从而保持最坏情况 O(n+m) 的时间。
4. Business – Database Aggregation for Sales Reporting | 商业——销售报表中的数据库聚合
A retail company’s SQL database contains tables Orders(OrderID, CustomerID, Date, Total) and OrderItems(OrderID, ProductID, Quantity, Price). Write a query to find the top three products by total revenue in March 2025, and discuss why an index on Date improves performance.
某零售公司的SQL数据库包含表 Orders(OrderID, CustomerID, Date, Total) 和 OrderItems(OrderID, ProductID, Quantity, Price)。写出一条查询，找出2025年3月总收入前三的产品，并讨论为何在Date上建立索引能提升性能。
The query joins both tables on OrderID, filters Date BETWEEN ‘2025-03-01’ AND ‘2025-03-31’, groups by ProductID, SUM(Quantity*Price) as revenue, and uses ORDER BY revenue DESC LIMIT 3. An index on Date allows the DBMS to perform an index range scan instead of a full table scan, drastically reducing disk I/O.
查询在 OrderID 上连接两表，用 Date BETWEEN ‘2025-03-01’ AND ‘2025-03-31’ 过滤，按 ProductID 分组，SUM(Quantity*Price) 求得收入，并用 ORDER BY revenue DESC LIMIT 3。在 Date 上建立索引可使DBMS执行索引范围扫描而非全表扫描，显著减少磁盘I/O。
5. Geography – Spatial Indexing with Quadtrees | 地理——四叉树空间索引
A GPS application stores millions of POIs (points of interest) on a 2D map. Explain how a quadtree partitions the space and how a range query (find all POIs within a rectangle) is processed more efficiently than with a linear list.
某GPS应用在二维地图上存储了数百万个兴趣点。解释四叉树如何划分空间，以及范围查询（找出矩形内所有兴趣点）为何比线性表更高效。
A quadtree recursively subdivides a square into four quadrants until each region contains ≤ k points. For a range query, the algorithm recursively visits only quadrants that intersect the query rectangle. This pruning reduces the average time to O(log n + m) for m results, compared with O(n) for a sequential scan.
四叉树递归地将正方形划分为四个象限，直到每个区域包含 ≤ k 个点。进行范围查询时，算法仅递归访问与查询矩形相交的象限。这种剪枝将平均时间降至 O(log n + m)（m 为结果数量），而顺序扫描需要 O(n)。
6. Design – Heuristic Evaluation of a Mobile Interface | 设计——移动界面的启发式评估
Using Nielsen’s heuristics, evaluate a shopping app where the checkout button is greyed out until a form is complete, but no error messages appear for empty fields. Identify two violated heuristics and propose a computing solution (e.g. real‑time validation feedback).
使用Nielsen启发式原则评估一款购物App：其结账按钮在表单未填完时呈灰色，但空白字段不显示任何错误信息。指出两个被违反的启发式原则，并提出一种计算机解决方案（如实时验证反馈）。
Violated heuristics include ‘Visibility of system status’ (users are not told why the button is disabled) and ‘Error prevention’ (no guidance prevents the error). A JavaScript validation module can highlight empty fields on blur and display clear hints, while keeping the button disabled until all constraints are met.
被违反的原则包括“系统状态的可见性”（用户不知道按钮为何禁用）和“错误预防”（没有指导来防止错误）。一个JavaScript验证模块可以在字段失焦时高亮空白字段并显示明确提示，同时保持按钮禁用直到所有约束满足。
7. Networking – Checksums and Data Integrity | 网络——校验和与数据完整性
A sensor transmits an 8‑bit data word 10110101 along with a 1‑byte checksum. The receiver calculates the checksum of the received data and compares it. Describe how an Internet checksum (one’s complement of one’s complement sum) detects single‑bit and burst errors, and compute the checksum for the given byte.
一个传感器传输8位数据字10110101和一个字节的校验和。接收方计算收到数据的校验和并进行比较。描述因特网校验和（反码和的反码）如何检测单比特和突发错误，并计算给定字节的校验和。
The Internet checksum splits data into 16‑bit words, sums them with carry wraparound, and takes the one’s complement. For a single byte, it is often padded. The one’s complement sum of 10110101 (0xB5) is 01001010 (0x4A). Its one’s complement is 10110101, yielding a checksum of 0xB5 if sent alone, but normally the header includes complemented 0x0000. It detects all 1‑bit errors and most burst errors.
因特网校验和将数据分为16位字，带进位回卷求和，再取反码。对于单个字节通常进行填充。10110101 (0xB5) 的反码和为 01001010 (0x4A)，其反码为 10110101，因此若单独发送校验和即为 0xB5，但首部通常包含取反后的 0x0000。它能检测所有单比特错误和大部分突发错误。
8. Law & Ethics – Data Protection and Encryption | 法律与伦理——数据保护与加密
Under the UK Data Protection Act 2018, a health‑tracking app must protect users’ medical data. Explain how AES‑256 encryption and salted hashing of passwords implement ‘appropriate technical measures’. Discuss the role of key management.
根据英国《2018年数据保护法》，一款健康追踪App必须保护用户的医疗数据。解释AES‑256加密和加盐哈希密码如何实现“适当的技术措施”，并讨论密钥管理的作用。
AES‑256 encrypts stored personal data so that a breach reveals only ciphertext. Salted hashing (e.g. bcrypt) prevents rainbow table attacks on passwords. Both are recognised as state‑of‑the‑art measures. Secure key management, such as storing keys in a hardware security module and enforcing strict access control, ensures that encryption cannot be bypassed.
AES‑256加密存储的个人数据，使得数据泄露只暴露密文。加盐哈希（如bcrypt）防止对密码的彩虹表攻击。两者都被视为最先进的技术措施。安全的密钥管理，如将密钥存储在硬件安全模块中并实施严格的访问控制，确保加密无法被绕过。
9. Machine Learning – Bias in Training Data | 机器学习——训练数据中的偏见
A hiring algorithm trained on historical CVs shows a preference for candidates from a small set of universities. From a computing perspective, explain how under‑representation in the training set leads to biased predictions, and suggest two algorithmic mitigation strategies.
某招聘算法基于历史简历训练后，表现出偏向来自少数几个大学的候选人。从计算角度解释训练集中代表性不足如何导致有偏见的预测，并提出两种算法缓解策略。
The model learns patterns that correlate university names with successful hires, treating the feature as a strong signal even when it reflects past human bias. Mitigation strategies include re‑sampling the dataset to balance representation and applying fairness constraints during training, such as adversarial debiasing or equalised odds post‑processing.
模型学到大学名称与录用成功之间的相关性，将这一特征视为强信号，即使它反映的是过去的人类偏见。缓解策略包括重采样数据集以平衡代表性，以及在训练过程中施加公平性约束，如对抗去偏或事后均衡几率处理。
10. Software Project Management – COCOMO Effort Estimation | 软件项目管理——COCOMO工作量估算
A development team estimates a project size of 50 KLOC using the basic COCOMO model. For an organic project, effort = 2.4 × (size)¹·⁰⁵ person‑months. Calculate the effort and determine the development time using the formula T = 2.5 × Effort⁰·³⁸. Discuss how this estimate informs sprint planning.
一个开发团队使用基本COCOMO模型估算项目规模为50 KLOC。对于有机式项目，工作量 = 2.4 × (规模)¹·⁰⁵ 人月。计算工作量并用公式 T = 2.5 × 工作量⁰·³⁸ 确定开发时间。讨论该估算如何为冲刺计划提供依据。
Effort = 2.4 × 50¹·⁰⁵ ≈ 2.4 × 60.2 ≈ 144.5 person‑months. T = 2.5 × 144.5⁰·³⁸ ≈ 2.5 × 6.62 ≈ 16.6 months. The estimate helps allocate total person‑months across sprints and validates whether the deadline is feasible. Agile teams use such macro‑estimates to set a realistic velocity baseline.
工作量 = 2.4 × 50¹·⁰⁵ ≈ 2.4 × 60.2 ≈ 144.5 人月。T = 2.5 × 144.5⁰·³⁸ ≈ 2.5 × 6.62 ≈ 16.6 个月。该估算有助于将总人月分配到各个冲刺，并验证截止日期是否可行。敏捷团队利用这类宏观估算设定现实的速度基线。
Published by TutorHao | Computer Science Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导