EE

Extended Essay, Computer Science topics

Students often fall into the trap of "building something" (which is an IA task). For the EE, they must "measure something."

Here is a curated list of IB Computer Science EE topics, categorized by domain and complexity.

🟢 Category 1: Algorithms & Data Structures (The "Safe" Zone)

These are excellent because they are highly controllable. Students can isolate variables easily and generate clear quantitative data (time, memory, operations).

  • Pathfinding Efficiency: "To what extent does the heuristic function (Manhattan vs. Euclidean distance) affect the execution time and path optimality of the A* algorithm in varying maze densities?"
    • Methodology: Generate random mazes of increasing density (10%, 20%... obstacles) and run both heuristics 100 times, averaging the time.
  • Sorting Data: "How does the degree of 'pre-sortedness' in a dataset affect the runtime efficiency of QuickSort compared to MergeSort?"
    • Methodology: Create datasets that are 0%, 25%, 50%, 75%, and 100% sorted. Run both algorithms and graph the performance degradation.
  • Data Structure Overhead: "To what extent does the use of a Hash Map versus a Binary Search Tree impact the retrieval time of user records as the dataset scales from $10^3$ to $10^6$ entries?"

🔵 Category 2: Artificial Intelligence & Machine Learning (High Interest)

Students love AI, but they must avoid "black box" usage. They need to manipulate the internal workings of the model.

  • Neural Network Architecture: "How does the number of hidden layers (1 vs. 3 vs. 5) in a Feed-Forward Neural Network impact the accuracy and training time when classifying handwritten digits (MNIST dataset)?"
    • Methodology: Train 3 separate models with identical learning rates/epochs but different depths. Compare the "cost" (time) vs. "benefit" (accuracy).
  • Image Compression via AI: "To what extent can a K-Means Clustering algorithm reduce the file size of .PNG images compared to standard lossy compression, while maintaining perceived visual quality?"
    • Methodology: Implement K-Means color quantization. Measure file size reduction vs. a standard metric like Peak Signal-to-Noise Ratio (PSNR).
  • Genetic Algorithms: "How does the mutation rate parameter affect the convergence speed of a Genetic Algorithm solving the Travelling Salesman Problem?"

🟠 Category 3: Cybersecurity & Encryption (Theoretical & Practical)

These require careful setup to ensure the student is measuring performance, not just describing how encryption works.

  • Encryption Overhead: "To what extent does key size (128-bit vs. 256-bit) in the AES algorithm affect battery consumption and CPU usage on mobile devices during file encryption?"
    • Methodology: Run a script on a laptop/phone that continuously encrypts files while logging CPU voltage/usage.
  • Password Strength Analysis: "To what extent does the inclusion of special characters versus increased length linearly or exponentially increase the time required for a Brute-force attack to crack a password (simulated)?"
    • Methodology: Write a simple brute-force script. Don't actually hack anything; just mathematically model or simulate the time to crack based on entropy calculations.

🔴 Category 4: Networking & Systems (Technical)

Good for students interested in hardware or back-end infrastructure.

  • Database Comparison: "To what extent does the read/write latency of a SQL (MySQL) database compare to a NoSQL (MongoDB) database when handling unstructured JSON data of increasing volume?"
  • Network Protocol Simulation: "How does the window size in the TCP sliding window protocol affect throughput and packet loss rates in a simulated high-latency network environment?"
    • Methodology: Use a network simulator (like NS-3 or Packet Tracer) or write a Python script to simulate packet transmission with varying window sizes.

⚠️ The "Danger Zone" (Topics to Avoid)

Unless they have a very unique angle:

  1. "Predicting the Stock Market": Almost always fails. The data is too noisy, and the student essentially concludes "it's random."
  2. "Building a Game/App": This is an Internal Assessment (IA), not an EE. If they write "How I built a Unity Game," they will score poorly. They must ask "How does the rendering technique in my game affect frame rate?"
  3. Pure History: "The History of the CPU" is a history essay, not a CS one.

💡 A Note on Artifacts for the EE

While the new curriculum emphasizes the paper, I strongly advise your students to build a "Lab Bench" artifact.

  • Example: If they are comparing Sorting Algorithms, their "artifact" is not a commercial sorting app, but a harness script that:
    1. Generates random arrays.
    2. Runs the sort.
    3. Times it with nanosecond precision.
    4. Exports the results to Excel/CSV.