Loading…

Worked solutions

IB Computer Science HL · Paper 2

Computer Science, Higher Level

Session: May 2014
Marks: 260
Duration: 1 hour 20 min
Calculator: Not allowed
Questions: 17

Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17

Keep your copy of the paper open alongside. We don't reproduce the questions: each part below is labelled to match the May 2014 Global paper, which your school or the IB can provide.

Practice mode

Hide our working until you've had a go.

Option A: Databases · Option

Questions 1–4 · 65 marks

Question 1

15 marks

Healthcare Regulatory Agency centralized product database and user interaction

Databases

(a)

3 marks

Define a database as a structured repository storing logically related data regarding a specific activity or domain.
Contrast this with an information system, which encompasses the underlying database as one core component.
Explain that an information system integrates the database with broader hardware, software, user interfaces, procedures, and personnel to collect, process, and distribute information.

Mentor tip

Always distinguish the storage layer (database) from the complete socio-technical ecosystem (information system).

(b)

2 marks

Guarantees data consistency across all participating pharmacies at any given moment.
Enables single-point central updates by the regulatory agency, immediately propagating changes without synchronizing local copies.

Mentor tip

Focus on single-source-of-truth advantages: real-time consistency and centralized maintenance.

(c)

1 mark

Identify the read-only query or retrieval function used to search and inspect authorized product specifications.

Mentor tip

Pharmacists only verify existing records; do not list write, update, or modification operations.

(d)

3 marks

Multiple pharmacists across different locations access the Healthcare Regulatory Agency database simultaneously.
The DBMS manages these overlapping requests concurrently so users experience seamless, parallel read access.
Equal, non-blocking access is preserved so one pharmacist's product query does not delay or corrupt another's session.

Mentor tip

Break concurrency down into: multiple users, simultaneous execution, and preserved integrity/access.

(e)

6 marks

Interaction Method 1 — Menu-driven / Form-based filtering:
1) Objective: Retrieve a filtered view of authorized medications targeting a specific condition (e.g., influenza).
2) Mechanism: The pharmacist navigates a graphical dropdown menu of therapeutic categories and selects 'influenza' to trigger a pre-configured lookup displaying matching drug specifications.
Interaction Method 2 — Parameterized Query By Example (QBE) / Search Form:
1) Objective: Locate suitable pediatric medications matching multiple clinical criteria (e.g., symptom = cough AND patient age < 12).
2) Mechanism: The user enters filter parameters directly into labeled form fields, which the interface translates into an underlying SQL SELECT query to return matching records.
Interaction Method 3 — Visual Report Generation:
1) Objective: Produce a formatted, printable summary of all authorized products manufactured by a specific pharmaceutical supplier.
2) Mechanism: The pharmacist clicks a report icon and selects a visual report template, prompting the system to query, group, and format the read-only dataset.

Mentor tip

For 6-mark discussion questions on user interaction, detail at least two distinct read-only scenarios covering what is requested, specific filter parameters, and the technical interface mechanism.

Question 2

10 marks

Relational database management system (RDBMS) architecture, schema, and data modelling

Databases

(a)(i)

1 mark

Define an RDBMS as specialized software that manages databases structured around logical tables (relations) linked by common attributes.

Mentor tip

Emphasize both the management software role and the relational (table-based) structure.

(a)(ii)

1 mark

Define a schema as the formal logical blueprint specifying the structure, tables, attributes, data types, and relationships of the entire database.

Mentor tip

Think of the schema as the structural definition rather than the actual data instances.

(b)

2 marks

1) Physical organization and layout of records on storage media (such as disk blocks, pages, and file structures).
2) Implementation of physical indexes and record ordering to optimize disk access speed.

Mentor tip

Physical level features always relate to low-level storage, indexing, and hardware allocation.

(c)

6 marks

1) Alignment with business objectives: Modelling clarifies system requirements early; without it, the resulting database may fail to support essential business workflows.
2) Accurate entity identification: Defining the core entities (tables) ensures all required domain objects are represented cleanly.
3) Attribute completeness and efficiency: Selecting necessary and sufficient attributes prevents missing data and avoids storage or processing inefficiencies.
4) Key and relationship specification: Establishing primary and foreign keys enables reliable record retrieval and complex multi-table joins while minimizing data duplication.

Mentor tip

Pair each data modelling feature (entities, attributes, keys, relationships) with the concrete consequence of omitting it.

Question 3

20 marks

OnlyScience scientific journal database: ERD, SQL queries, views, and 3NF normalization

Databases

(a)

3 marks

Identify the entities and cardinality constraints:
- OnlyScience (1) ----< publishes >---- (N) Journal
- Editor (1) ----< is responsible for >---- (N) Journal
Combine into a labeled Entity-Relationship Diagram showing 1-to-many relationships from Publisher to Journal and from Editor to Journal.

Mentor tip

Clearly label both the relationship verbs ('publishes', 'responsible for') and the 1:N cardinalities.

(b)(i)

1 mark

A primary key is an attribute or composite group of attributes that uniquely identifies every individual record (tuple) within a relation.

Mentor tip

Include the keyword 'unique' in any primary key definition.

(b)(ii)

1 mark

A foreign key is an attribute in one relation that references the primary key of another relation to establish a logical link between the tables.

Mentor tip

State clearly that it corresponds to the primary key of another table.

(c)

1 mark

Locate the row in the 'activity' table where pers-id equals 'Boris' and extract the corresponding journal-id:
Result: Astrophysics

Mentor tip

Return only the exact attribute value requested in the SELECT clause.

(d)

5 marks

Identify the required projection: person.phone.
Identify the three tables to join: person (p), activity (a), and journal (j).
Specify the join conditions and filter predicate:
1) p.name = a.pers-id (links person to activity)
2) a.journal-id = j.title (links activity to journal)
3) j.number = 'Special' (filters for the target journal issue)
Construct the complete SQL statement:
SELECT p.phone FROM person p, activity a, journal j WHERE p.name = a.pers-id AND a.journal-id = j.title AND j.number = 'Special';

Mentor tip

When joining three tables, ensure you have two join equality conditions plus the specific filtering condition.

(e)

3 marks

A query applies vertical filtering (selecting specific columns) and horizontal filtering (applying WHERE conditions such as role = 'editor').
It dynamically extracts and displays only the relevant subset of data tailored to the publisher's requirements.
Sensitive or unneeded rows (such as reviewer activities) remain hidden from the user's virtual table.

Mentor tip

Explain how column projection and row selection combine to present a tailored subset of the database.

(f)

3 marks

A standard view is part of the external/logical schema rather than the physical schema.
Therefore, no separate physical table containing the view's rows is permanently stored on disk.
Instead, the DBMS stores only the underlying query definition (macro) and dynamically recomputes the virtual result set from base tables whenever invoked.

Mentor tip

Distinguish between storing the query definition (metadata) and storing duplicated physical rows.

(g)

3 marks

Analyze dependencies in 'submission': paper-id has multiple reviewers (repeating group), and editor determines journal-title and number (transitive dependency).
Decompose into three 3NF relations with clear Primary Keys (PK) and Foreign Keys (FK):
1) SUBMISSION (paper-id [PK], editor [FK])
2) EDITORS (editor [PK], journal-title, number)
3) REVIEWERS (paper-id [PK, FK], reviewer [PK])

Mentor tip

Ensure every non-key attribute depends solely on the primary key of its respective table.

Question 4

20 marks

Data warehousing architecture, cluster analysis in data mining, time-variance, and e-commerce privacy

Databases

(a)

2 marks

A data warehouse is a centralized, subject-oriented repository that integrates historical and current data extracted from multiple disparate sources.
It is structured and optimized specifically to support analytical reporting, trend analysis, and strategic decision-making over long time periods.

Mentor tip

Highlight three pillars: integrated from multiple sources, accumulated over long time horizons, and built for analytics.

(b)

4 marks

1) Purpose and Operations: An operational database (OLTP) is designed for high-frequency day-to-day read/write transactions, whereas a data warehouse (OLAP) is designed for complex read-only analytical queries and large-scale reporting.
2) Normalization and Time-Horizon: Operational databases are highly normalized to eliminate redundancy and reflect only the current state, whereas data warehouses use denormalized multidimensional schemas (star/snowflake) storing historical, time-variant snapshots to maximize query speed.

Mentor tip

Contrast operational DBs and data warehouses across two clear axes, such as transaction processing vs analytics, and current state vs historical data.

(c)(i)

1 mark

Data mining is the automated computational process of discovering hidden patterns, correlations, and actionable knowledge within massive datasets.

Mentor tip

Focus on automated pattern discovery across large volumes of data.

(c)(ii)

3 marks

Cluster analysis partitions records into distinct groups (clusters) based on attribute similarity without predefined category labels.
Items within the same cluster share high similarity (e.g., customers with comparable purchasing behavior), while items in different clusters are dissimilar.
Analysts inspect these emergent clusters to uncover previously unseen market segments or behavioral structures.

Mentor tip

Emphasize grouping by similarity to reveal previously unknown natural structures.

(d)(i)

3 marks

Real-world business metrics change continuously over time, so warehouse records must preserve historical states rather than overwriting them.
Valid longitudinal analysis and forecasting require comparing current performance against historical baselines.
As new periodic snapshots arrive, analytical models recompute trends across the chronological sequence.

Mentor tip

Explain that preserving historical snapshots is essential for longitudinal trend and forecasting analysis.

(d)(ii)

2 marks

Include an explicit timestamp or unique Time_ID foreign key in every fact table record loaded into the warehouse.
Link these records to a dedicated Time dimension table so queries can filter and aggregate metrics by day, quarter, or year.

Mentor tip

Mention stamping every loaded record with a timestamp or Time-ID attribute.

(e)

5 marks

1) Invasive profiling: E-commerce platforms combine factual identity details (address, payment info) with browsing logs and purchase histories via persistent tracking cookies and pixel tags.
2) Price discrimination and exclusion: Cross-referencing behavioral profiles allows retailers to offer dynamic pricing (charging wealthier profiles more) or deny services to users flagged as credit risks.
3) Third-party sharing and corporate acquisitions: User profiles are frequently shared with third-party advertising networks or transferred during corporate buyouts, exposing consumers to secondary data breaches or altered privacy policies without explicit consent.

Mentor tip

Discuss how tracking mechanisms (cookies), behavioral cross-referencing (price discrimination), and third-party data sharing threaten consumer privacy.

Option B: Modelling and simulation · Option

Questions 5–8 · 65 marks

Question 5

16 marks

National debt and GDP percentage mathematical model and spreadsheet simulation

Modelling and simulation

(a)

4 marks

Define the discrete time step i representing each 3-month quarter over multiple years.
Specify the required periodic inputs: Government Spending (GS_i), Income (I_i), Gross Domestic Product (GDP_i), and initial/previous debt (CD_{i-1}).
Formulate the recurrence equation for current debt: CD_i = GS_i + CD_{i-1} - I_i.
Compute the relative debt ratio: %Debt_i = (CD_i / GDP_i) * 100, and store the resulting tuple (i, GS_i, I_i, CD_i, GDP_i, %Debt_i) as a time-series row.

C D_{i} = G S_{i} + C D_{i - 1} - I_{i} % Debt_{i} = (\frac{C D _{i}}{G D P _{i}}) \times 100

Mentor tip

Explicitly state inputs, the two mathematical formulas, and how each quarter's output feeds into the next period.

(b)

6 marks

1) Select a spreadsheet application (e.g., Microsoft Excel) and organize rows by quarterly periods (Quarter 1, Quarter 2, ...).
2) Insert dedicated columns for Tax Revenue (T_i) and Bond Sales (B_i).
3) Define the Income column using a formula summing those components: I_i = T_i + B_i (e.g., =B2+C2).
4) Reference the previous row's Current Debt cell to compute the new Current Debt: CD_i = GS_i + CD_{i-1} - I_i.
5) Replicate (fill down) the formulas across multiple future rows to project multi-year trajectories.
6) Perform 'what-if' scenario testing by varying tax rates or bond issuance figures and plotting a line chart to visualize long-term debt trends.

I_{i} = Taxes_{i} + Bonds_{i}

Mentor tip

Reference concrete spreadsheet mechanisms: columns, cell formulas, fill-down replication, and parameter tweaking.

(c)

6 marks

Advantages to governments:
1) Long-term foresight: Policymakers can project multi-year fiscal impacts of current spending or tax decisions rather than focusing only on the immediate budget cycle.
2) Early warning and iterative policy testing: Enables proactive intervention and safe experimentation with different fiscal rules until sustainable debt ratios are found.
Limitations and risks:
1) Unforeseen shocks: Historical trends and static assumptions cannot predict sudden economic crises, wars, or pandemics, which can render projections misleading.
2) Model rigidity: Over-reliance on simplified equations may neglect complex socio-economic feedback loops and discourage creative human policy initiatives.

Mentor tip

Balance your discussion with both strategic policy benefits and the inherent limitations of simplified economic assumptions.

Question 6

14 marks

Children's 3D arithmetic learning game variables, rendering, and socio-technical impact

Modelling and simulation

(a)

4 marks

1) Operand variables (e.g., num1, num2: integer) and Operator variable (op: char/string): Randomly regenerated at the start of each new question.
2) ExpectedAnswer (numeric): Recalculated and stored each round by applying the chosen operator to the two random numbers.
3) PlayerGuess (numeric): Updated each round via keyboard/controller input from the child.
4) Score (integer): Initialized to 0 at the start of the session and incremented by 1 whenever PlayerGuess matches ExpectedAnswer.

Mentor tip

You must state both the variable's role AND how its value changes during each question cycle.

(b)

2 marks

Characters (monsters, wizards) are stored internally as abstract 3D mathematical wireframes and surface properties.
Rendering is required to apply shading, textures, lighting, and perspective projection so that each character movement appears as a fluid, lifelike 3D animation rather than flat geometry.

Mentor tip

Connect mathematical 3D models to visual realism (lighting, textures, fluid frame-by-frame motion).

(c)

8 marks

Technical Implications:
1) Computational load: Real-time 3D rendering of animated characters and castle mazes requires high-speed CPUs/GPUs and substantial RAM; otherwise frame rates drop and gameplay lags.
2) Hardware accessibility: High system requirements may exclude children from lower-income households or schools using older computers lacking dedicated graphics cards.
Social and Educational Implications:
1) Enhanced engagement: Immersive 3D visuals and gamified rewards make arithmetic exciting, motivating reluctant learners to practice mental math more frequently.
2) Psychological and behavioral risks: Scary dungeon animals triggered by wrong answers may frighten young children, causing math anxiety, or conversely, excessive gamification may foster screen addiction over genuine learning.

Mentor tip

Structure 8-mark discussions clearly into two elaborated technical points and two elaborated social points.

Question 7

15 marks

Urban traffic congestion data collection, simulation modelling, and evaluation

Modelling and simulation

(a)

4 marks

Deploy automated road sensors (inductive loops, infrared counters) and traffic cameras at strategic bottlenecks such as intersections, roundabouts, and traffic lights.
Record vehicle counts, speeds, and queue lengths continuously across all hours of the day (capturing peak rush hours and off-peak baselines).
Extend data collection over an extended calendar duration to capture variations across weekdays, weekends, school holidays, and adverse weather conditions.

Mentor tip

Automated sensors and cameras over varied time periods are essential; manual surveys or questionnaires cannot capture real-time traffic flow.

(b)

4 marks

Represent the city road network as a mathematical graph where nodes model junctions (traffic lights, roundabouts) and weighted edges model road segments with distances, lane capacities, and direction rules.
Convert the empirical sensor data into time-dependent vehicle arrival streams (probabilities and flow rates).
Step the simulation clock forward through a virtual day or week, routing vehicle agents through the network according to junction control rules to observe queue formation.

Mentor tip

Distinguish how static infrastructure (graph nodes/edges) and dynamic traffic (time-stream arrival rates) are combined.

(c)

4 marks

Criterion 1 — Average cross-city transit time during peak rush hour:
- Assessment: Measure the simulated travel duration for vehicles commuting from suburbs to the city center under each of the four road changes to see which minimizes delay.
Criterion 2 — Queue length and throughput at key bottlenecks (or localized emissions):
- Assessment: Compare maximum waiting queues at major junctions before and after rule modifications to verify that congestion is genuinely reduced rather than simply shifted to adjacent streets.

Mentor tip

State a measurable metric and explain how comparing its simulated values guides decision-making.

(d)

3 marks

Novel human behavioral shifts—such as how many drivers will actually switch to cycling or public transit when new lanes are built—lack historical baseline sensor data.
Estimating these modal-shift rates relies on public surveys or speculative assumptions, which are often unreliable or overly optimistic.
Feeding inaccurate behavioral assumptions into the simulation can significantly distort predicted traffic reductions.

Mentor tip

Focus on the lack of historical empirical data and the unpredictability of human commuter behavior.

Question 8

20 marks

Natural language processing challenges and neural network character recognition

Modelling and simulation

(a)

6 marks

1) Lexical Ambiguity (Vocabulary): In human languages, a single word often carries multiple meanings depending on context (polysemy, e.g., 'bank' or 'foot'), whereas machine languages assign a single, unambiguous meaning to every token.
2) Syntactic Flexibility vs Rigid Grammar: Human communication frequently bends grammar rules, uses idioms, or omits words while remaining understandable, whereas machine languages enforce strict, inflexible syntax rules.
3) Semantic and Contextual Nuance: In human language, literal word-for-word translation sorted by syntax often misses sarcasm, tone, or implied context, whereas machine instructions have deterministic, context-free semantics.

Mentor tip

Structure your 6-mark contrast around the three linguistic levels: vocabulary (ambiguity), syntax (flexibility vs strictness), and semantics (context).

(b)

6 marks

1) Real-time Cross-Lingual Translation: Instantaneous, accurate translation of web pages, scientific publications, and live speech removes international language barriers.
2) Intuitive Human-Computer Interaction: Voice-controlled assistants and hands-free robotic control empower users (including those with physical disabilities) to operate complex systems using natural speech.
3) Automated Knowledge Extraction and AI Support: Intelligent conversational agents, clinical diagnostic assistants, and automated summarizers can digest vast text repositories to answer complex domain questions.

Mentor tip

State three distinct application areas (translation, voice control/accessibility, intelligent assistants) and explain the concrete benefit of each.

(c)

2 marks

Construct a directed multi-layer diagram ordered left-to-right:
1) Input Layer (3 nodes: I1, I2, I3) -> Hidden Layer (e.g., H1..Hn) -> Output Layer (O1..Om).
2) Draw directed arrows connecting every input node forward to the hidden layer nodes, and from the hidden layer nodes forward to the output layer.

Mentor tip

Ensure layers are ordered Input -> Hidden -> Output and links show clear forward directionality.

(d)

6 marks

1) Feature Input: Present numerous varied handwritten samples of each character to the input layer nodes (e.g., pixel intensities or stroke features).
2) Forward Propagation: Signals pass from the input nodes through weighted connections into the hidden layer, which computes intermediate activation probabilities mapped toward target ASCII character codes at the output layer.
3) Error Evaluation: Compare the network's output probabilities against the known target ASCII code for the training sample.
4) Weight Adjustment (Backpropagation / Iterative Selection): Retain and reinforce connection weights that yield high probability for the correct character while penalizing incorrect pathways.
5) Convergence: Repeat across thousands of training epochs until recognition accuracy exceeds a required threshold.
6) Inference: Once trained, an unseen handwritten character is fed into the input layer and the network outputs the most probable ASCII character.

Mentor tip

Walk chronologically through training: varied inputs -> hidden layer weight propagation -> comparing output to target ASCII -> adjusting weights -> testing unseen handwriting.

Option C: Web science · Option

Questions 9–13 · 65 marks

Question 9

15 marks

HTTP/HTTPS protocols, client-side JavaScript, and server-side scripting in e-commerce

Web science

(a)

2 marks

The client web browser formats and transmits an HTTP request message targeting a specific resource on the web server.
The web server processes the request and returns an HTTP response message containing the status code and the requested HTML/media content.

Mentor tip

Describe the fundamental Request-Response cycle clearly between client and server.

(b)

4 marks

1) Server Authentication: Uses digital SSL/TLS certificates verified by trusted Certificate Authorities to confirm the website's genuine identity, preventing man-in-the-middle spoofing.
2) Data Encryption: Encrypts all transmitted packets between browser and server so that intercepted traffic cannot be deciphered or read by eavesdroppers.

Mentor tip

State both security mechanisms (authentication + encryption) and the specific threat each mitigates.

(c)(i)

1 mark

The browser fetches and executes the external JavaScript file named 'anyFile.js' locally on the client machine.

Mentor tip

Identify both the file name and that it is executed by the client browser.

(c)(ii)

2 marks

1) Performs immediate client-side validation of user form inputs before submission to the server.
2) Enables dynamic DOM manipulation, interactive animations, or asynchronous content updates without reloading the entire web page.

Mentor tip

Highlight interactive client-side enhancements such as input validation, animations, or asynchronous updates.

(d)

6 marks

1) Server-side execution: Scripts written in languages like PHP, Python, or Node.js execute on the web server before plain HTML is returned to the client browser.
2) Database integration for transactions: Server-side scripts securely connect to backend inventory and payment databases to verify stock, process orders, and record purchases without exposing credentials to the client.
3) Personalization and targeted offers: By querying stored customer profiles and purchase histories after login/cookie recognition, the script dynamically generates customized HTML pages featuring tailored promotional offers.

Mentor tip

Link server-side execution to both secure database transactions (purchasing) and dynamic content customization (personalized offers).

Question 10

15 marks

Web crawlers, the deep web, hubs and authorities, and SEO ethics

Web science

(a)

3 marks

1) The crawler begins with a seed list of initial URLs to visit.
2) It fetches each page, parses the HTML to extract all outgoing hyperlinks, and appends unvisited URLs to its crawl frontier.
3) It recursively visits those discovered links breadth-first or depth-first until a stopping condition or depth limit is reached.

Mentor tip

Mention the initial seed URLs, link extraction, and recursive/iterative traversal.

(b)(i)

1 mark

Deep web pages lack static incoming hyperlinks from the surface web, or require authentication/form submission, preventing crawlers from discovering them.

Mentor tip

Focus on the absence of static hypertext links or protection behind login/query forms.

(b)(ii)

1 mark

They are dynamically generated on-the-fly by server-side scripts (e.g., PHP) querying a backend database in response to user input.

Mentor tip

Connect deep web pages to dynamic database queries.

(c)

4 marks

Define a hub as a directory or resource page that contains many outgoing links pointing to relevant pages.
Define an authority as a high-value content page that receives many incoming links from reputable hubs.
A hub's score increases if it points to strong authorities, and an authority's score increases if it is pointed to by strong hubs.
The search engine iteratively computes both scores (as in the HITS algorithm) to rank overall page importance.

Mentor tip

Clearly distinguish outgoing links (hubs) from incoming links (authorities) and explain their mutually reinforcing relationship.

(d)

6 marks

Moral responsibilities of Search Engine Designers:
- Maintain algorithmic neutrality so organic rankings reflect genuine relevance rather than hidden commercial bias.
- Clearly label paid advertisements or sponsored links so users are not deceived into thinking they are organic top results.
Moral responsibilities of Website Developers:
- Employ ethical ('white-hat') Search Engine Optimization (SEO), such as accurate meta tags, clean site architecture, and relevant keywords that genuinely reflect page content.
- Refrain from deceptive ('black-hat') spamdexing practices—such as keyword stuffing, hidden text, cloaking, or link farms—that manipulate rankings and mislead users.

Mentor tip

Address both stakeholders (search engine designers and web developers) and contrast ethical practices with deceptive practices.

Question 11

15 marks

Ubiquitous, mobile, and grid computing, intellectual property, and the democratic web

Web science

(a)

2 marks

Define ubiquitous computing as the seamless integration of microprocessors and network-connected sensors into everyday physical objects and environments.
Example: Smart RFID-tagged grocery items or wearable biometric health monitors woven into clothing.

Mentor tip

Pair a clear definition of 'everywhere embedded computing' with a concrete physical example.

(b)

1 mark

Limited battery life/power capacity (or wireless bandwidth constraints, signal coverage dead zones, and heightened interception risks).

Mentor tip

Any single valid constraint—battery life, bandwidth, coverage, or security—earns the mark.

(c)

4 marks

1) Cost-effectiveness vs Capital Expenditure: Grid computing harnesses thousands of inexpensive networked desktops (or volunteer nodes), avoiding the massive purchase and maintenance cost of a dedicated supercomputer.
2) Network Latency and Reliability: Grid nodes communicate over slower internet links compared to a supercomputer's ultra-fast internal bus, and individual volunteer machines may disconnect unpredictably or pose data security risks.

Mentor tip

Examine two distinct factors (e.g., cost, network speed, security, hardware availability) with a clear explanation for each.

(d)

2 marks

Intellectual property (IP) is the broad intangible creation of the human mind (such as original articles, music, software, or designs) belonging to its creator.
Copyright is the specific legal protection mechanism granting the IP owner exclusive rights to control reproduction and distribution of that work for a designated period.

Mentor tip

Distinguish the creation/ownership concept (IP) from the legal enforcement right (copyright).

(e)

6 marks

Concept of the Democratic Web:
- A neutral, open web ensures every website is equally accessible regardless of the owner's financial power or geographic location (net neutrality).
Contrasting Online Retailers Example:
- An independent artisan or small startup in a developing country can host an online storefront accessible on equal technical footing with a multinational giant like Amazon.
Consequences if Web Democracy is Lost:
- ISPs could create a two-tier internet ('fast lanes' for wealthy corporations), throttling traffic to small competitors.
- Dominant corporations and wealthier nations would monopolize online visibility, stifling small business innovation and harming developing economies.

Mentor tip

Cover all three required elements: define the democratic web, give a contrasting retailer example, and explain the socio-economic consequences of losing neutrality.

Question 12

11 marks

University web graph metrics, directed connectivity, and text vs image search

Web science

(a)(i)

1 mark

524,970 represents the total number of distinct web pages (nodes/vertices) within the Stansfield university web graph.

Mentor tip

In a web graph, nodes correspond to individual web pages.

(a)(ii)

1 mark

6,345,270 represents the total number of directed hyperlinks (edges) connecting those web pages.

Mentor tip

In a web graph, edges correspond to hyperlinks between pages.

(b)

1 mark

The subset of web pages and internal links belonging to a single academic department or faculty (e.g., the Computer Science department sub-site).

Mentor tip

Give a concrete university-related subset of pages, such as a specific faculty or department.

(c)(i)

2 marks

The edges connecting the nodes have arrowheads.
These arrows specify the one-way direction of navigation from the source page containing the hyperlink to the target page.

Mentor tip

Identify the visual feature (arrows) and its meaning (direction of hyperlink traversal).

(c)(ii)

2 marks

A graph is strongly connected only if a directed path exists from every node to every other node.
In the given diagram, Node A has no incoming edges (in-degree 0), making it impossible to navigate back to Node A from any other node (B, C, D, E, or F).

Mentor tip

Define strong connectivity first, then point out a specific unreachable node (Node A has only outgoing links).

(d)

4 marks

Text-based image search:
- Matches user keywords against textual metadata surrounding the image (filenames, alt-tags, captions).
- Fast to execute, but quality is limited by human subjectivity, missing tags, or inaccurate descriptions.
Content/Image-based search (Reverse Image Search):
- Analyzes visual features of a query image (color histograms, textures, shapes, pattern/facial recognition) to find visually similar matches.
- Delivers more accurate visual matches independent of language, but requires intensive server-side computation.

Mentor tip

Contrast how each method works (metadata matching vs visual feature extraction) and their trade-offs (tag subjectivity vs computational cost).

Question 13

9 marks

Folksonomies in the semantic web and ambient intelligence in smart homes

Web science

(a)

3 marks

Folksonomies arise when everyday users collaboratively tag web resources (photos, articles, bookmarks) with descriptive keywords.
Over time, popular tags converge into shared bottom-up vocabularies that add semantic metadata structure to unstructured content.
This crowd-sourced categorization helps search engines and semantic web tools interpret page meaning more effectively.

Mentor tip

Link user-generated tagging to emergent shared vocabularies and improved machine interpretability.

(b)

6 marks

Concept of Ambient Intelligence (AmI):
- An environment embedded with networked sensors, processors, and actuators that invisibly detects human presence and adapts automatically to individual preferences.
Identification and Sensing Technology:
- Uses biometric sensors (facial recognition cameras, voiceprint microphones, iris scanners) or wearable RFID/Bluetooth tokens to distinguish the mother from the daughter upon entry.
Differentiated Automated Responses:
- When the mother is identified, the hallway display shows her calendar reminders, family location updates, and triggers the kitchen smart kettle.
- When the daughter is identified, the home automation controller adjusts her bedroom thermostat, turns on her study lighting, and streams her preferred music playlist.

Mentor tip

Structure your 6-mark response into three 2-mark blocks: define Ambient Intelligence, specify the identification hardware (sensors/biometrics), and contrast specific automated actions for mother vs daughter.

Option D: Object-oriented programming · Option

Questions 14–17 · 65 marks

Question 14

13 marks

Zoo taxonomy object model: inheritance, association, UML, and polymorphism

Object-oriented programming

(a)

1 mark

Inheritance ('is-a' relationship): The Species class is a subclass that extends (inherits from) the Genus superclass.

Mentor tip

Look at the class header 'public class Species extends Genus' to identify inheritance immediately.

(b)

1 mark

Aggregation / Association ('has-a' relationship): The Specimen class contains an instance variable (toa) that holds a reference to a Species object.

Mentor tip

When one class stores an instance of another class as a field, the relationship is 'has-a' (aggregation/association).

(c)

4 marks

Draw a standard 3-compartment UML class box:
1) Top compartment (Class Name): Species (or Species extends Genus)
2) Middle compartment (Attributes): - speciesName : String
3) Bottom compartment (Methods): + Species(s : String, g : String), + setSpeciesName(s : String) : void, + getSpeciesName() : String, + toString() : String, + equals(s : Species) : boolean

Mentor tip

Include visibility modifiers (- for private attributes, + for public methods) and return types in the 3-tier box.

(d)

4 marks

1) Code reuse and reduced duplication: Because Species inherits from Genus and Specimen references a shared Species object, common genus/species attributes and methods are written once rather than duplicated across every animal.
2) Simplified testing and maintenance: Inherited methods already verified in the Genus superclass do not need to be re-written, and updates to a Species object's details automatically reflect across all associated Specimen instances.

Mentor tip

State two distinct software engineering benefits (e.g., code reuse, easier testing/maintenance, real-world modularity) and elaborate on each.

(e)(i)

2 marks

Each toString() method is scoped to its respective class definition in the hierarchy.
Because 'human' is instantiated as a Species object, the compiler/runtime binds the call to the overridden toString() method defined in the Species subclass rather than the Genus superclass.

Mentor tip

Explain that method resolution selects the implementation belonging to the specific subclass type.

(e)(ii)

1 mark

Polymorphism (or method overriding).

Mentor tip

Either 'polymorphism' or 'method overriding' is the standard OOP term here.

Question 15

14 marks

Encapsulation, accessor methods, Genus class implementation, and subclass design trade-offs

Object-oriented programming

(a)

1 mark

Encapsulation is the bundling of data (instance variables) and the methods that operate on that data into a single class unit while restricting direct external access to the internal representation.

Mentor tip

Mention both bundling data/methods together and hiding internal data behind public methods.

(b)

4 marks

1) Data protection and integrity: Marking instance variables private prevents external classes from accidentally corrupting object state; controlled setter methods can enforce validation.
2) Modularity and ease of maintenance: Internal data structures or implementations inside a class can be modified without breaking external code, provided the public method signatures remain unchanged.

Mentor tip

Provide two clear advantages (such as data security and maintainability) with a full sentence explaining each.

(c)

1 mark

getName() (or getCage() or getTOA()).

Mentor tip

Accessor methods (getters) retrieve private attribute values.

(d)

1 mark

name (or cageNumber or toa).

Mentor tip

Instance variables are the private fields declared at the top of the Specimen class.

(e)

3 marks

Declare the Genus class with a private String attribute genusName:
public class Genus {
private String genusName;
public Genus(String g) { setGenusName(g); }
public void setGenusName(String g) { genusName = g; }
public String getGenusName() { return genusName; }
public String toString() { return "Genus: " + genusName; }
}

Mentor tip

Ensure your code includes all three requested elements: single-parameter String constructor, getter/setter methods, and toString().

(f)

4 marks

Advantage:
- Direct inheritance of members: A Specimen object would directly inherit all methods and attributes of Species, allowing direct method calls without chaining through a separate 'toa' object reference.
Disadvantage:
- Redundancy and logical inconsistency: Every individual animal instance would store its own duplicated copy of species-level data (wasting memory), and updating a species name would require updating every individual specimen of that species.

Mentor tip

Contrast the convenience of direct method access against memory duplication across many individual specimens.

Question 16

18 marks

Modifying Specimen for markings and constructing array search/listing algorithms

Object-oriented programming

(a)

4 marks

1) Recognize that physical markings belong to an individual animal rather than an entire species.
2) Add a private instance variable (e.g., private String markings;) to the Specimen class and update its constructor to initialize it.
3) Add public accessor and mutator methods (getMarkings() and setMarkings(String m)) in Specimen.
4) Update the Specimen.toString() method to include the markings description in its returned string.

Mentor tip

Trace the impact of adding an attribute across the whole class: field declaration, constructor, getter/setter, and toString().

(b)

8 marks

Construct the method to iterate over the array, compare Species objects via .equals(), and print the total:
public void countSpecimens(Specimen[] animals, Species s) {
int count = 0;
for (int i = 0; i < animals.length; i++) {
if (animals[i] != null && s.equals(animals[i].getTOA())) {
count++;
}
}
System.out.println(count);
}

Mentor tip

Use animals[i].getTOA() to retrieve each specimen's Species object, and compare using the Species .equals() method rather than ==.

(c)

6 marks

Initialize an empty collection or auxiliary array UNIQUE to track distinct Species encountered.
Loop through every Specimen element in the input array ANIMALS:
1) Extract CURRENT_SPECIES = ANIMALS[i].getTOA().
2) Search UNIQUE to check whether CURRENT_SPECIES is already present.
3) If not found in UNIQUE, add CURRENT_SPECIES to UNIQUE and output CURRENT_SPECIES.toString().
Pseudocode:
UNIQUE = new Collection()
loop i from 0 to ANIMALS.length - 1
FOUND = false
UNIQUE.resetNext()
loop while UNIQUE.hasNext()
if UNIQUE.getNext().equals(ANIMALS[i].getTOA()) then
FOUND = true
end if
end loop
if NOT FOUND then
UNIQUE.addItem(ANIMALS[i].getTOA())
output ANIMALS[i].getTOA().toString()
end if
end loop

Mentor tip

To list distinct species without duplicates, always maintain a seen-list and check membership before adding/printing.

Question 17

20 marks

JETS LinkedList ADT operations for managing Specimen and unique Species collections

Object-oriented programming

(a)

2 marks

1) Implementation details (how data is physically stored and manipulated in memory) are hidden from the user.
2) It is defined by a standardized public interface of supported operations/methods (such as addHead, getNext, isEmpty).

Mentor tip

Remember the two core features of an ADT: hidden internal representation and a standard set of operations.

(b)

4 marks

Instantiate a new LinkedList, iterate through the Specimen array, insert each element, and return the list:
LinkedList makeList(Specimen[] animals) {
LinkedList list = new LinkedList();
for (int i = 0; i < animals.length; i++) {
if (animals[i] != null) {
list.addTail(animals[i]); // or addHead(animals[i])
}
}
return list;
}

Mentor tip

Ensure you instantiate the LinkedList, loop to animals.length, call addHead/addTail, and return the populated list.

(c)

6 marks

Create a new LinkedList for Species, traverse the input animals list safely using getHead() and getNext(), extract each specimen's Species via getTOA(), and add it:
public LinkedList makeSpeciesList(LinkedList animals) {
LinkedList speciesList = new LinkedList();
Specimen current = (Specimen) animals.getHead();
while (current != null) {
speciesList.addTail(current.getTOA());
current = (Specimen) animals.getNext();
}
return speciesList;
}

Mentor tip

Use a while (current != null) loop starting from animals.getHead() so that an empty input list is handled safely without errors.

(d)

8 marks

Traverse allSpecies with an outer loop, check membership in uniqueSpecies with an inner loop using .equals(), and insert only unseen Species:
public LinkedList makeSpeciesListUnique(LinkedList allSpecies) {
LinkedList uniqueSpecies = new LinkedList();
Species currentSpecies = (Species) allSpecies.getHead();
while (currentSpecies != null) {
boolean found = false;
Species existing = (Species) uniqueSpecies.getHead();
while (existing != null) {
if (existing.equals(currentSpecies)) {
found = true;
}
existing = (Species) uniqueSpecies.getNext();
}
if (!found) {
uniqueSpecies.addTail(currentSpecies);
}
currentSpecies = (Species) allSpecies.getNext();
}
return uniqueSpecies;
}

Mentor tip

Reset the boolean flag (found = false) at the start of every outer loop iteration before scanning the uniqueSpecies list.

Our analysis

Where the marks are

Each part sorted by the topic it mostly tests. Databases carries the most marks on this paper.

Databases65 marks · 25%
Q1, Q2, Q3, Q4
Modelling and simulation65 marks · 25%
Q5, Q6, Q7, Q8
Web science65 marks · 25%
Q9, Q10, Q11, Q12, Q13
Object-oriented programming65 marks · 25%
Q14, Q15, Q16, Q17

Mentor insights

Traps to watch for

01
Databases: OnlyScience scientific journal database: ERD, SQL queries, views, · Q3
Clearly label both the relationship verbs ('publishes', 'responsible for') and the 1:N cardinalities.
02
Modelling and simulation: Natural language processing challenges and neural network charact · Q8
Structure your 6-mark contrast around the three linguistic levels: vocabulary (ambiguity), syntax (flexibility vs strictness), and semantics (context).
03
Web science: HTTP/HTTPS protocols, client-side JavaScript, and server-side scr · Q9
Describe the fundamental Request-Response cycle clearly between client and server.
04
Object-oriented programming: JETS LinkedList ADT operations for managing Specimen and unique S · Q17
Remember the two core features of an ADT: hidden internal representation and a standard set of operations.

Solutions and tips are GeorgeTheTutor's own; the questions are not reproduced here. Not affiliated with or endorsed by the International Baccalaureate Organization.

Want someone to go through your paper with you?

Our maths mentors took these exams recently. One-to-one, they'll find where your marks are going and fix it before May.

Find your maths mentor More past papers

Worked solutions

IB Computer Science HL · Paper 2

Computer Science, Higher Level

Session: May 2014
Marks: 260
Duration: 1 hour 20 min
Calculator: Not allowed
Questions: 17

Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17

Keep your copy of the paper open alongside. We don't reproduce the questions: each part below is labelled to match the May 2014 Global paper, which your school or the IB can provide.

Practice mode

Hide our working until you've had a go.

Option A: Databases · Option

Questions 1–4 · 65 marks

Question 1

15 marks

Healthcare Regulatory Agency centralized product database and user interaction

Databases

(a)

3 marks

Define a database as a structured repository storing logically related data regarding a specific activity or domain.
Contrast this with an information system, which encompasses the underlying database as one core component.
Explain that an information system integrates the database with broader hardware, software, user interfaces, procedures, and personnel to collect, process, and distribute information.

Mentor tip

Always distinguish the storage layer (database) from the complete socio-technical ecosystem (information system).

(b)

2 marks

Guarantees data consistency across all participating pharmacies at any given moment.
Enables single-point central updates by the regulatory agency, immediately propagating changes without synchronizing local copies.

Mentor tip

Focus on single-source-of-truth advantages: real-time consistency and centralized maintenance.

(c)

1 mark

Identify the read-only query or retrieval function used to search and inspect authorized product specifications.

Mentor tip

Pharmacists only verify existing records; do not list write, update, or modification operations.

(d)

3 marks

Multiple pharmacists across different locations access the Healthcare Regulatory Agency database simultaneously.
The DBMS manages these overlapping requests concurrently so users experience seamless, parallel read access.
Equal, non-blocking access is preserved so one pharmacist's product query does not delay or corrupt another's session.

Mentor tip

Break concurrency down into: multiple users, simultaneous execution, and preserved integrity/access.

(e)

6 marks

Interaction Method 1 — Menu-driven / Form-based filtering:
1) Objective: Retrieve a filtered view of authorized medications targeting a specific condition (e.g., influenza).
2) Mechanism: The pharmacist navigates a graphical dropdown menu of therapeutic categories and selects 'influenza' to trigger a pre-configured lookup displaying matching drug specifications.
Interaction Method 2 — Parameterized Query By Example (QBE) / Search Form:
1) Objective: Locate suitable pediatric medications matching multiple clinical criteria (e.g., symptom = cough AND patient age < 12).
2) Mechanism: The user enters filter parameters directly into labeled form fields, which the interface translates into an underlying SQL SELECT query to return matching records.
Interaction Method 3 — Visual Report Generation:
1) Objective: Produce a formatted, printable summary of all authorized products manufactured by a specific pharmaceutical supplier.
2) Mechanism: The pharmacist clicks a report icon and selects a visual report template, prompting the system to query, group, and format the read-only dataset.

Mentor tip

For 6-mark discussion questions on user interaction, detail at least two distinct read-only scenarios covering what is requested, specific filter parameters, and the technical interface mechanism.

Question 2

10 marks

Relational database management system (RDBMS) architecture, schema, and data modelling

Databases

(a)(i)

1 mark

Define an RDBMS as specialized software that manages databases structured around logical tables (relations) linked by common attributes.

Mentor tip

Emphasize both the management software role and the relational (table-based) structure.

(a)(ii)

1 mark

Define a schema as the formal logical blueprint specifying the structure, tables, attributes, data types, and relationships of the entire database.

Mentor tip

Think of the schema as the structural definition rather than the actual data instances.

(b)

2 marks

1) Physical organization and layout of records on storage media (such as disk blocks, pages, and file structures).
2) Implementation of physical indexes and record ordering to optimize disk access speed.

Mentor tip

Physical level features always relate to low-level storage, indexing, and hardware allocation.

(c)

6 marks

1) Alignment with business objectives: Modelling clarifies system requirements early; without it, the resulting database may fail to support essential business workflows.
2) Accurate entity identification: Defining the core entities (tables) ensures all required domain objects are represented cleanly.
3) Attribute completeness and efficiency: Selecting necessary and sufficient attributes prevents missing data and avoids storage or processing inefficiencies.
4) Key and relationship specification: Establishing primary and foreign keys enables reliable record retrieval and complex multi-table joins while minimizing data duplication.

Mentor tip

Pair each data modelling feature (entities, attributes, keys, relationships) with the concrete consequence of omitting it.

Question 3

20 marks

OnlyScience scientific journal database: ERD, SQL queries, views, and 3NF normalization

Databases

(a)

3 marks

Identify the entities and cardinality constraints:
- OnlyScience (1) ----< publishes >---- (N) Journal
- Editor (1) ----< is responsible for >---- (N) Journal
Combine into a labeled Entity-Relationship Diagram showing 1-to-many relationships from Publisher to Journal and from Editor to Journal.

Mentor tip

Clearly label both the relationship verbs ('publishes', 'responsible for') and the 1:N cardinalities.

(b)(i)

1 mark

A primary key is an attribute or composite group of attributes that uniquely identifies every individual record (tuple) within a relation.

Mentor tip

Include the keyword 'unique' in any primary key definition.

(b)(ii)

1 mark

A foreign key is an attribute in one relation that references the primary key of another relation to establish a logical link between the tables.

Mentor tip

State clearly that it corresponds to the primary key of another table.

(c)

1 mark

Locate the row in the 'activity' table where pers-id equals 'Boris' and extract the corresponding journal-id:
Result: Astrophysics

Mentor tip

Return only the exact attribute value requested in the SELECT clause.

(d)

5 marks

Identify the required projection: person.phone.
Identify the three tables to join: person (p), activity (a), and journal (j).
Specify the join conditions and filter predicate:
1) p.name = a.pers-id (links person to activity)
2) a.journal-id = j.title (links activity to journal)
3) j.number = 'Special' (filters for the target journal issue)
Construct the complete SQL statement:
SELECT p.phone FROM person p, activity a, journal j WHERE p.name = a.pers-id AND a.journal-id = j.title AND j.number = 'Special';

Mentor tip

When joining three tables, ensure you have two join equality conditions plus the specific filtering condition.

(e)

3 marks

A query applies vertical filtering (selecting specific columns) and horizontal filtering (applying WHERE conditions such as role = 'editor').
It dynamically extracts and displays only the relevant subset of data tailored to the publisher's requirements.
Sensitive or unneeded rows (such as reviewer activities) remain hidden from the user's virtual table.

Mentor tip

Explain how column projection and row selection combine to present a tailored subset of the database.

(f)

3 marks

A standard view is part of the external/logical schema rather than the physical schema.
Therefore, no separate physical table containing the view's rows is permanently stored on disk.
Instead, the DBMS stores only the underlying query definition (macro) and dynamically recomputes the virtual result set from base tables whenever invoked.

Mentor tip

Distinguish between storing the query definition (metadata) and storing duplicated physical rows.

(g)

3 marks

Analyze dependencies in 'submission': paper-id has multiple reviewers (repeating group), and editor determines journal-title and number (transitive dependency).
Decompose into three 3NF relations with clear Primary Keys (PK) and Foreign Keys (FK):
1) SUBMISSION (paper-id [PK], editor [FK])
2) EDITORS (editor [PK], journal-title, number)
3) REVIEWERS (paper-id [PK, FK], reviewer [PK])

Mentor tip

Ensure every non-key attribute depends solely on the primary key of its respective table.

Question 4

20 marks

Data warehousing architecture, cluster analysis in data mining, time-variance, and e-commerce privacy

Databases

(a)

2 marks

A data warehouse is a centralized, subject-oriented repository that integrates historical and current data extracted from multiple disparate sources.
It is structured and optimized specifically to support analytical reporting, trend analysis, and strategic decision-making over long time periods.

Mentor tip

Highlight three pillars: integrated from multiple sources, accumulated over long time horizons, and built for analytics.

(b)

4 marks

1) Purpose and Operations: An operational database (OLTP) is designed for high-frequency day-to-day read/write transactions, whereas a data warehouse (OLAP) is designed for complex read-only analytical queries and large-scale reporting.
2) Normalization and Time-Horizon: Operational databases are highly normalized to eliminate redundancy and reflect only the current state, whereas data warehouses use denormalized multidimensional schemas (star/snowflake) storing historical, time-variant snapshots to maximize query speed.

Mentor tip

Contrast operational DBs and data warehouses across two clear axes, such as transaction processing vs analytics, and current state vs historical data.

(c)(i)

1 mark

Data mining is the automated computational process of discovering hidden patterns, correlations, and actionable knowledge within massive datasets.

Mentor tip

Focus on automated pattern discovery across large volumes of data.

(c)(ii)

3 marks

Cluster analysis partitions records into distinct groups (clusters) based on attribute similarity without predefined category labels.
Items within the same cluster share high similarity (e.g., customers with comparable purchasing behavior), while items in different clusters are dissimilar.
Analysts inspect these emergent clusters to uncover previously unseen market segments or behavioral structures.

Mentor tip

Emphasize grouping by similarity to reveal previously unknown natural structures.

(d)(i)

3 marks

Real-world business metrics change continuously over time, so warehouse records must preserve historical states rather than overwriting them.
Valid longitudinal analysis and forecasting require comparing current performance against historical baselines.
As new periodic snapshots arrive, analytical models recompute trends across the chronological sequence.

Mentor tip

Explain that preserving historical snapshots is essential for longitudinal trend and forecasting analysis.

(d)(ii)

2 marks

Include an explicit timestamp or unique Time_ID foreign key in every fact table record loaded into the warehouse.
Link these records to a dedicated Time dimension table so queries can filter and aggregate metrics by day, quarter, or year.

Mentor tip

Mention stamping every loaded record with a timestamp or Time-ID attribute.

(e)

5 marks

1) Invasive profiling: E-commerce platforms combine factual identity details (address, payment info) with browsing logs and purchase histories via persistent tracking cookies and pixel tags.
2) Price discrimination and exclusion: Cross-referencing behavioral profiles allows retailers to offer dynamic pricing (charging wealthier profiles more) or deny services to users flagged as credit risks.
3) Third-party sharing and corporate acquisitions: User profiles are frequently shared with third-party advertising networks or transferred during corporate buyouts, exposing consumers to secondary data breaches or altered privacy policies without explicit consent.

Mentor tip

Discuss how tracking mechanisms (cookies), behavioral cross-referencing (price discrimination), and third-party data sharing threaten consumer privacy.

Option B: Modelling and simulation · Option

Questions 5–8 · 65 marks

Question 5

16 marks

National debt and GDP percentage mathematical model and spreadsheet simulation

Modelling and simulation

(a)

4 marks

Define the discrete time step i representing each 3-month quarter over multiple years.
Specify the required periodic inputs: Government Spending (GS_i), Income (I_i), Gross Domestic Product (GDP_i), and initial/previous debt (CD_{i-1}).
Formulate the recurrence equation for current debt: CD_i = GS_i + CD_{i-1} - I_i.
Compute the relative debt ratio: %Debt_i = (CD_i / GDP_i) * 100, and store the resulting tuple (i, GS_i, I_i, CD_i, GDP_i, %Debt_i) as a time-series row.

C D_{i} = G S_{i} + C D_{i - 1} - I_{i} % Debt_{i} = (\frac{C D _{i}}{G D P _{i}}) \times 100

Mentor tip

Explicitly state inputs, the two mathematical formulas, and how each quarter's output feeds into the next period.

(b)

6 marks

1) Select a spreadsheet application (e.g., Microsoft Excel) and organize rows by quarterly periods (Quarter 1, Quarter 2, ...).
2) Insert dedicated columns for Tax Revenue (T_i) and Bond Sales (B_i).
3) Define the Income column using a formula summing those components: I_i = T_i + B_i (e.g., =B2+C2).
4) Reference the previous row's Current Debt cell to compute the new Current Debt: CD_i = GS_i + CD_{i-1} - I_i.
5) Replicate (fill down) the formulas across multiple future rows to project multi-year trajectories.
6) Perform 'what-if' scenario testing by varying tax rates or bond issuance figures and plotting a line chart to visualize long-term debt trends.

I_{i} = Taxes_{i} + Bonds_{i}

Mentor tip

Reference concrete spreadsheet mechanisms: columns, cell formulas, fill-down replication, and parameter tweaking.

(c)

6 marks

Advantages to governments:
1) Long-term foresight: Policymakers can project multi-year fiscal impacts of current spending or tax decisions rather than focusing only on the immediate budget cycle.
2) Early warning and iterative policy testing: Enables proactive intervention and safe experimentation with different fiscal rules until sustainable debt ratios are found.
Limitations and risks:
1) Unforeseen shocks: Historical trends and static assumptions cannot predict sudden economic crises, wars, or pandemics, which can render projections misleading.
2) Model rigidity: Over-reliance on simplified equations may neglect complex socio-economic feedback loops and discourage creative human policy initiatives.

Mentor tip

Balance your discussion with both strategic policy benefits and the inherent limitations of simplified economic assumptions.

Question 6

14 marks

Children's 3D arithmetic learning game variables, rendering, and socio-technical impact

Modelling and simulation

(a)

4 marks

1) Operand variables (e.g., num1, num2: integer) and Operator variable (op: char/string): Randomly regenerated at the start of each new question.
2) ExpectedAnswer (numeric): Recalculated and stored each round by applying the chosen operator to the two random numbers.
3) PlayerGuess (numeric): Updated each round via keyboard/controller input from the child.
4) Score (integer): Initialized to 0 at the start of the session and incremented by 1 whenever PlayerGuess matches ExpectedAnswer.

Mentor tip

You must state both the variable's role AND how its value changes during each question cycle.

(b)

2 marks

Characters (monsters, wizards) are stored internally as abstract 3D mathematical wireframes and surface properties.
Rendering is required to apply shading, textures, lighting, and perspective projection so that each character movement appears as a fluid, lifelike 3D animation rather than flat geometry.

Mentor tip

Connect mathematical 3D models to visual realism (lighting, textures, fluid frame-by-frame motion).

(c)

8 marks

Technical Implications:
1) Computational load: Real-time 3D rendering of animated characters and castle mazes requires high-speed CPUs/GPUs and substantial RAM; otherwise frame rates drop and gameplay lags.
2) Hardware accessibility: High system requirements may exclude children from lower-income households or schools using older computers lacking dedicated graphics cards.
Social and Educational Implications:
1) Enhanced engagement: Immersive 3D visuals and gamified rewards make arithmetic exciting, motivating reluctant learners to practice mental math more frequently.
2) Psychological and behavioral risks: Scary dungeon animals triggered by wrong answers may frighten young children, causing math anxiety, or conversely, excessive gamification may foster screen addiction over genuine learning.

Mentor tip

Structure 8-mark discussions clearly into two elaborated technical points and two elaborated social points.

Question 7

15 marks

Urban traffic congestion data collection, simulation modelling, and evaluation

Modelling and simulation

(a)

4 marks

Deploy automated road sensors (inductive loops, infrared counters) and traffic cameras at strategic bottlenecks such as intersections, roundabouts, and traffic lights.
Record vehicle counts, speeds, and queue lengths continuously across all hours of the day (capturing peak rush hours and off-peak baselines).
Extend data collection over an extended calendar duration to capture variations across weekdays, weekends, school holidays, and adverse weather conditions.

Mentor tip

Automated sensors and cameras over varied time periods are essential; manual surveys or questionnaires cannot capture real-time traffic flow.

(b)

4 marks

Represent the city road network as a mathematical graph where nodes model junctions (traffic lights, roundabouts) and weighted edges model road segments with distances, lane capacities, and direction rules.
Convert the empirical sensor data into time-dependent vehicle arrival streams (probabilities and flow rates).
Step the simulation clock forward through a virtual day or week, routing vehicle agents through the network according to junction control rules to observe queue formation.

Mentor tip

Distinguish how static infrastructure (graph nodes/edges) and dynamic traffic (time-stream arrival rates) are combined.

(c)

4 marks

Criterion 1 — Average cross-city transit time during peak rush hour:
- Assessment: Measure the simulated travel duration for vehicles commuting from suburbs to the city center under each of the four road changes to see which minimizes delay.
Criterion 2 — Queue length and throughput at key bottlenecks (or localized emissions):
- Assessment: Compare maximum waiting queues at major junctions before and after rule modifications to verify that congestion is genuinely reduced rather than simply shifted to adjacent streets.

Mentor tip

State a measurable metric and explain how comparing its simulated values guides decision-making.

(d)

3 marks

Novel human behavioral shifts—such as how many drivers will actually switch to cycling or public transit when new lanes are built—lack historical baseline sensor data.
Estimating these modal-shift rates relies on public surveys or speculative assumptions, which are often unreliable or overly optimistic.
Feeding inaccurate behavioral assumptions into the simulation can significantly distort predicted traffic reductions.

Mentor tip

Focus on the lack of historical empirical data and the unpredictability of human commuter behavior.

Question 8

20 marks

Natural language processing challenges and neural network character recognition

Modelling and simulation

(a)

6 marks

1) Lexical Ambiguity (Vocabulary): In human languages, a single word often carries multiple meanings depending on context (polysemy, e.g., 'bank' or 'foot'), whereas machine languages assign a single, unambiguous meaning to every token.
2) Syntactic Flexibility vs Rigid Grammar: Human communication frequently bends grammar rules, uses idioms, or omits words while remaining understandable, whereas machine languages enforce strict, inflexible syntax rules.
3) Semantic and Contextual Nuance: In human language, literal word-for-word translation sorted by syntax often misses sarcasm, tone, or implied context, whereas machine instructions have deterministic, context-free semantics.

Mentor tip

Structure your 6-mark contrast around the three linguistic levels: vocabulary (ambiguity), syntax (flexibility vs strictness), and semantics (context).

(b)

6 marks

1) Real-time Cross-Lingual Translation: Instantaneous, accurate translation of web pages, scientific publications, and live speech removes international language barriers.
2) Intuitive Human-Computer Interaction: Voice-controlled assistants and hands-free robotic control empower users (including those with physical disabilities) to operate complex systems using natural speech.
3) Automated Knowledge Extraction and AI Support: Intelligent conversational agents, clinical diagnostic assistants, and automated summarizers can digest vast text repositories to answer complex domain questions.

Mentor tip

State three distinct application areas (translation, voice control/accessibility, intelligent assistants) and explain the concrete benefit of each.

(c)

2 marks

Construct a directed multi-layer diagram ordered left-to-right:
1) Input Layer (3 nodes: I1, I2, I3) -> Hidden Layer (e.g., H1..Hn) -> Output Layer (O1..Om).
2) Draw directed arrows connecting every input node forward to the hidden layer nodes, and from the hidden layer nodes forward to the output layer.

Mentor tip

Ensure layers are ordered Input -> Hidden -> Output and links show clear forward directionality.

(d)

6 marks

1) Feature Input: Present numerous varied handwritten samples of each character to the input layer nodes (e.g., pixel intensities or stroke features).
2) Forward Propagation: Signals pass from the input nodes through weighted connections into the hidden layer, which computes intermediate activation probabilities mapped toward target ASCII character codes at the output layer.
3) Error Evaluation: Compare the network's output probabilities against the known target ASCII code for the training sample.
4) Weight Adjustment (Backpropagation / Iterative Selection): Retain and reinforce connection weights that yield high probability for the correct character while penalizing incorrect pathways.
5) Convergence: Repeat across thousands of training epochs until recognition accuracy exceeds a required threshold.
6) Inference: Once trained, an unseen handwritten character is fed into the input layer and the network outputs the most probable ASCII character.

Mentor tip

Walk chronologically through training: varied inputs -> hidden layer weight propagation -> comparing output to target ASCII -> adjusting weights -> testing unseen handwriting.

Option C: Web science · Option

Questions 9–13 · 65 marks

Question 9

15 marks

HTTP/HTTPS protocols, client-side JavaScript, and server-side scripting in e-commerce

Web science

(a)

2 marks

The client web browser formats and transmits an HTTP request message targeting a specific resource on the web server.
The web server processes the request and returns an HTTP response message containing the status code and the requested HTML/media content.

Mentor tip

Describe the fundamental Request-Response cycle clearly between client and server.

(b)

4 marks

1) Server Authentication: Uses digital SSL/TLS certificates verified by trusted Certificate Authorities to confirm the website's genuine identity, preventing man-in-the-middle spoofing.
2) Data Encryption: Encrypts all transmitted packets between browser and server so that intercepted traffic cannot be deciphered or read by eavesdroppers.

Mentor tip

State both security mechanisms (authentication + encryption) and the specific threat each mitigates.

(c)(i)

1 mark

The browser fetches and executes the external JavaScript file named 'anyFile.js' locally on the client machine.

Mentor tip

Identify both the file name and that it is executed by the client browser.

(c)(ii)

2 marks

1) Performs immediate client-side validation of user form inputs before submission to the server.
2) Enables dynamic DOM manipulation, interactive animations, or asynchronous content updates without reloading the entire web page.

Mentor tip

Highlight interactive client-side enhancements such as input validation, animations, or asynchronous updates.

(d)

6 marks

1) Server-side execution: Scripts written in languages like PHP, Python, or Node.js execute on the web server before plain HTML is returned to the client browser.
2) Database integration for transactions: Server-side scripts securely connect to backend inventory and payment databases to verify stock, process orders, and record purchases without exposing credentials to the client.
3) Personalization and targeted offers: By querying stored customer profiles and purchase histories after login/cookie recognition, the script dynamically generates customized HTML pages featuring tailored promotional offers.

Mentor tip

Link server-side execution to both secure database transactions (purchasing) and dynamic content customization (personalized offers).

Question 10

15 marks

Web crawlers, the deep web, hubs and authorities, and SEO ethics

Web science

(a)

3 marks

1) The crawler begins with a seed list of initial URLs to visit.
2) It fetches each page, parses the HTML to extract all outgoing hyperlinks, and appends unvisited URLs to its crawl frontier.
3) It recursively visits those discovered links breadth-first or depth-first until a stopping condition or depth limit is reached.

Mentor tip

Mention the initial seed URLs, link extraction, and recursive/iterative traversal.

(b)(i)

1 mark

Deep web pages lack static incoming hyperlinks from the surface web, or require authentication/form submission, preventing crawlers from discovering them.

Mentor tip

Focus on the absence of static hypertext links or protection behind login/query forms.

(b)(ii)

1 mark

They are dynamically generated on-the-fly by server-side scripts (e.g., PHP) querying a backend database in response to user input.

Mentor tip

Connect deep web pages to dynamic database queries.

(c)

4 marks

Define a hub as a directory or resource page that contains many outgoing links pointing to relevant pages.
Define an authority as a high-value content page that receives many incoming links from reputable hubs.
A hub's score increases if it points to strong authorities, and an authority's score increases if it is pointed to by strong hubs.
The search engine iteratively computes both scores (as in the HITS algorithm) to rank overall page importance.

Mentor tip

Clearly distinguish outgoing links (hubs) from incoming links (authorities) and explain their mutually reinforcing relationship.

(d)

6 marks

Moral responsibilities of Search Engine Designers:
- Maintain algorithmic neutrality so organic rankings reflect genuine relevance rather than hidden commercial bias.
- Clearly label paid advertisements or sponsored links so users are not deceived into thinking they are organic top results.
Moral responsibilities of Website Developers:
- Employ ethical ('white-hat') Search Engine Optimization (SEO), such as accurate meta tags, clean site architecture, and relevant keywords that genuinely reflect page content.
- Refrain from deceptive ('black-hat') spamdexing practices—such as keyword stuffing, hidden text, cloaking, or link farms—that manipulate rankings and mislead users.

Mentor tip

Address both stakeholders (search engine designers and web developers) and contrast ethical practices with deceptive practices.

Question 11

15 marks

Ubiquitous, mobile, and grid computing, intellectual property, and the democratic web

Web science

(a)

2 marks

Define ubiquitous computing as the seamless integration of microprocessors and network-connected sensors into everyday physical objects and environments.
Example: Smart RFID-tagged grocery items or wearable biometric health monitors woven into clothing.

Mentor tip

Pair a clear definition of 'everywhere embedded computing' with a concrete physical example.

(b)

1 mark

Limited battery life/power capacity (or wireless bandwidth constraints, signal coverage dead zones, and heightened interception risks).

Mentor tip

Any single valid constraint—battery life, bandwidth, coverage, or security—earns the mark.

(c)

4 marks

1) Cost-effectiveness vs Capital Expenditure: Grid computing harnesses thousands of inexpensive networked desktops (or volunteer nodes), avoiding the massive purchase and maintenance cost of a dedicated supercomputer.
2) Network Latency and Reliability: Grid nodes communicate over slower internet links compared to a supercomputer's ultra-fast internal bus, and individual volunteer machines may disconnect unpredictably or pose data security risks.

Mentor tip

Examine two distinct factors (e.g., cost, network speed, security, hardware availability) with a clear explanation for each.

(d)

2 marks

Intellectual property (IP) is the broad intangible creation of the human mind (such as original articles, music, software, or designs) belonging to its creator.
Copyright is the specific legal protection mechanism granting the IP owner exclusive rights to control reproduction and distribution of that work for a designated period.

Mentor tip

Distinguish the creation/ownership concept (IP) from the legal enforcement right (copyright).

(e)

6 marks

Concept of the Democratic Web:
- A neutral, open web ensures every website is equally accessible regardless of the owner's financial power or geographic location (net neutrality).
Contrasting Online Retailers Example:
- An independent artisan or small startup in a developing country can host an online storefront accessible on equal technical footing with a multinational giant like Amazon.
Consequences if Web Democracy is Lost:
- ISPs could create a two-tier internet ('fast lanes' for wealthy corporations), throttling traffic to small competitors.
- Dominant corporations and wealthier nations would monopolize online visibility, stifling small business innovation and harming developing economies.

Mentor tip

Cover all three required elements: define the democratic web, give a contrasting retailer example, and explain the socio-economic consequences of losing neutrality.

Question 12

11 marks

University web graph metrics, directed connectivity, and text vs image search

Web science

(a)(i)

1 mark

524,970 represents the total number of distinct web pages (nodes/vertices) within the Stansfield university web graph.

Mentor tip

In a web graph, nodes correspond to individual web pages.

(a)(ii)

1 mark

6,345,270 represents the total number of directed hyperlinks (edges) connecting those web pages.

Mentor tip

In a web graph, edges correspond to hyperlinks between pages.

(b)

1 mark

The subset of web pages and internal links belonging to a single academic department or faculty (e.g., the Computer Science department sub-site).

Mentor tip

Give a concrete university-related subset of pages, such as a specific faculty or department.

(c)(i)

2 marks

The edges connecting the nodes have arrowheads.
These arrows specify the one-way direction of navigation from the source page containing the hyperlink to the target page.

Mentor tip

Identify the visual feature (arrows) and its meaning (direction of hyperlink traversal).

(c)(ii)

2 marks

A graph is strongly connected only if a directed path exists from every node to every other node.
In the given diagram, Node A has no incoming edges (in-degree 0), making it impossible to navigate back to Node A from any other node (B, C, D, E, or F).

Mentor tip

Define strong connectivity first, then point out a specific unreachable node (Node A has only outgoing links).

(d)

4 marks

Text-based image search:
- Matches user keywords against textual metadata surrounding the image (filenames, alt-tags, captions).
- Fast to execute, but quality is limited by human subjectivity, missing tags, or inaccurate descriptions.
Content/Image-based search (Reverse Image Search):
- Analyzes visual features of a query image (color histograms, textures, shapes, pattern/facial recognition) to find visually similar matches.
- Delivers more accurate visual matches independent of language, but requires intensive server-side computation.

Mentor tip

Contrast how each method works (metadata matching vs visual feature extraction) and their trade-offs (tag subjectivity vs computational cost).

Question 13

9 marks

Folksonomies in the semantic web and ambient intelligence in smart homes

Web science

(a)

3 marks

Folksonomies arise when everyday users collaboratively tag web resources (photos, articles, bookmarks) with descriptive keywords.
Over time, popular tags converge into shared bottom-up vocabularies that add semantic metadata structure to unstructured content.
This crowd-sourced categorization helps search engines and semantic web tools interpret page meaning more effectively.

Mentor tip

Link user-generated tagging to emergent shared vocabularies and improved machine interpretability.

(b)

6 marks

Concept of Ambient Intelligence (AmI):
- An environment embedded with networked sensors, processors, and actuators that invisibly detects human presence and adapts automatically to individual preferences.
Identification and Sensing Technology:
- Uses biometric sensors (facial recognition cameras, voiceprint microphones, iris scanners) or wearable RFID/Bluetooth tokens to distinguish the mother from the daughter upon entry.
Differentiated Automated Responses:
- When the mother is identified, the hallway display shows her calendar reminders, family location updates, and triggers the kitchen smart kettle.
- When the daughter is identified, the home automation controller adjusts her bedroom thermostat, turns on her study lighting, and streams her preferred music playlist.

Mentor tip

Option D: Object-oriented programming · Option

Questions 14–17 · 65 marks

Question 14

13 marks

Zoo taxonomy object model: inheritance, association, UML, and polymorphism

Object-oriented programming

(a)

1 mark

Inheritance ('is-a' relationship): The Species class is a subclass that extends (inherits from) the Genus superclass.

Mentor tip

Look at the class header 'public class Species extends Genus' to identify inheritance immediately.

(b)

1 mark

Aggregation / Association ('has-a' relationship): The Specimen class contains an instance variable (toa) that holds a reference to a Species object.

Mentor tip

When one class stores an instance of another class as a field, the relationship is 'has-a' (aggregation/association).

(c)

4 marks

Draw a standard 3-compartment UML class box:
1) Top compartment (Class Name): Species (or Species extends Genus)
2) Middle compartment (Attributes): - speciesName : String
3) Bottom compartment (Methods): + Species(s : String, g : String), + setSpeciesName(s : String) : void, + getSpeciesName() : String, + toString() : String, + equals(s : Species) : boolean

Mentor tip

Include visibility modifiers (- for private attributes, + for public methods) and return types in the 3-tier box.

(d)

4 marks

1) Code reuse and reduced duplication: Because Species inherits from Genus and Specimen references a shared Species object, common genus/species attributes and methods are written once rather than duplicated across every animal.
2) Simplified testing and maintenance: Inherited methods already verified in the Genus superclass do not need to be re-written, and updates to a Species object's details automatically reflect across all associated Specimen instances.

Mentor tip

State two distinct software engineering benefits (e.g., code reuse, easier testing/maintenance, real-world modularity) and elaborate on each.

(e)(i)

2 marks

Each toString() method is scoped to its respective class definition in the hierarchy.
Because 'human' is instantiated as a Species object, the compiler/runtime binds the call to the overridden toString() method defined in the Species subclass rather than the Genus superclass.

Mentor tip

Explain that method resolution selects the implementation belonging to the specific subclass type.

(e)(ii)

1 mark

Polymorphism (or method overriding).

Mentor tip

Either 'polymorphism' or 'method overriding' is the standard OOP term here.

Question 15

14 marks

Encapsulation, accessor methods, Genus class implementation, and subclass design trade-offs

Object-oriented programming

(a)

1 mark

Encapsulation is the bundling of data (instance variables) and the methods that operate on that data into a single class unit while restricting direct external access to the internal representation.

Mentor tip

Mention both bundling data/methods together and hiding internal data behind public methods.

(b)

4 marks

1) Data protection and integrity: Marking instance variables private prevents external classes from accidentally corrupting object state; controlled setter methods can enforce validation.
2) Modularity and ease of maintenance: Internal data structures or implementations inside a class can be modified without breaking external code, provided the public method signatures remain unchanged.

Mentor tip

Provide two clear advantages (such as data security and maintainability) with a full sentence explaining each.

(c)

1 mark

getName() (or getCage() or getTOA()).

Mentor tip

Accessor methods (getters) retrieve private attribute values.

(d)

1 mark

name (or cageNumber or toa).

Mentor tip

Instance variables are the private fields declared at the top of the Specimen class.

(e)

3 marks

Declare the Genus class with a private String attribute genusName:
public class Genus {
private String genusName;
public Genus(String g) { setGenusName(g); }
public void setGenusName(String g) { genusName = g; }
public String getGenusName() { return genusName; }
public String toString() { return "Genus: " + genusName; }
}

Mentor tip

Ensure your code includes all three requested elements: single-parameter String constructor, getter/setter methods, and toString().

(f)

4 marks

Advantage:
- Direct inheritance of members: A Specimen object would directly inherit all methods and attributes of Species, allowing direct method calls without chaining through a separate 'toa' object reference.
Disadvantage:
- Redundancy and logical inconsistency: Every individual animal instance would store its own duplicated copy of species-level data (wasting memory), and updating a species name would require updating every individual specimen of that species.

Mentor tip

Contrast the convenience of direct method access against memory duplication across many individual specimens.

Question 16

18 marks

Modifying Specimen for markings and constructing array search/listing algorithms

Object-oriented programming

(a)

4 marks

1) Recognize that physical markings belong to an individual animal rather than an entire species.
2) Add a private instance variable (e.g., private String markings;) to the Specimen class and update its constructor to initialize it.
3) Add public accessor and mutator methods (getMarkings() and setMarkings(String m)) in Specimen.
4) Update the Specimen.toString() method to include the markings description in its returned string.

Mentor tip

Trace the impact of adding an attribute across the whole class: field declaration, constructor, getter/setter, and toString().

(b)

8 marks

Construct the method to iterate over the array, compare Species objects via .equals(), and print the total:
public void countSpecimens(Specimen[] animals, Species s) {
int count = 0;
for (int i = 0; i < animals.length; i++) {
if (animals[i] != null && s.equals(animals[i].getTOA())) {
count++;
}
}
System.out.println(count);
}

Mentor tip

Use animals[i].getTOA() to retrieve each specimen's Species object, and compare using the Species .equals() method rather than ==.

(c)

6 marks

Initialize an empty collection or auxiliary array UNIQUE to track distinct Species encountered.
Loop through every Specimen element in the input array ANIMALS:
1) Extract CURRENT_SPECIES = ANIMALS[i].getTOA().
2) Search UNIQUE to check whether CURRENT_SPECIES is already present.
3) If not found in UNIQUE, add CURRENT_SPECIES to UNIQUE and output CURRENT_SPECIES.toString().
Pseudocode:
UNIQUE = new Collection()
loop i from 0 to ANIMALS.length - 1
FOUND = false
UNIQUE.resetNext()
loop while UNIQUE.hasNext()
if UNIQUE.getNext().equals(ANIMALS[i].getTOA()) then
FOUND = true
end if
end loop
if NOT FOUND then
UNIQUE.addItem(ANIMALS[i].getTOA())
output ANIMALS[i].getTOA().toString()
end if
end loop

Mentor tip

To list distinct species without duplicates, always maintain a seen-list and check membership before adding/printing.

Question 17

20 marks

JETS LinkedList ADT operations for managing Specimen and unique Species collections

Object-oriented programming

(a)

2 marks

1) Implementation details (how data is physically stored and manipulated in memory) are hidden from the user.
2) It is defined by a standardized public interface of supported operations/methods (such as addHead, getNext, isEmpty).

Mentor tip

Remember the two core features of an ADT: hidden internal representation and a standard set of operations.

(b)

4 marks

Instantiate a new LinkedList, iterate through the Specimen array, insert each element, and return the list:
LinkedList makeList(Specimen[] animals) {
LinkedList list = new LinkedList();
for (int i = 0; i < animals.length; i++) {
if (animals[i] != null) {
list.addTail(animals[i]); // or addHead(animals[i])
}
}
return list;
}

Mentor tip

Ensure you instantiate the LinkedList, loop to animals.length, call addHead/addTail, and return the populated list.

(c)

6 marks

Create a new LinkedList for Species, traverse the input animals list safely using getHead() and getNext(), extract each specimen's Species via getTOA(), and add it:
public LinkedList makeSpeciesList(LinkedList animals) {
LinkedList speciesList = new LinkedList();
Specimen current = (Specimen) animals.getHead();
while (current != null) {
speciesList.addTail(current.getTOA());
current = (Specimen) animals.getNext();
}
return speciesList;
}

Mentor tip

Use a while (current != null) loop starting from animals.getHead() so that an empty input list is handled safely without errors.

(d)

8 marks

Traverse allSpecies with an outer loop, check membership in uniqueSpecies with an inner loop using .equals(), and insert only unseen Species:
public LinkedList makeSpeciesListUnique(LinkedList allSpecies) {
LinkedList uniqueSpecies = new LinkedList();
Species currentSpecies = (Species) allSpecies.getHead();
while (currentSpecies != null) {
boolean found = false;
Species existing = (Species) uniqueSpecies.getHead();
while (existing != null) {
if (existing.equals(currentSpecies)) {
found = true;
}
existing = (Species) uniqueSpecies.getNext();
}
if (!found) {
uniqueSpecies.addTail(currentSpecies);
}
currentSpecies = (Species) allSpecies.getNext();
}
return uniqueSpecies;
}

Mentor tip

Reset the boolean flag (found = false) at the start of every outer loop iteration before scanning the uniqueSpecies list.

Our analysis

Where the marks are

Each part sorted by the topic it mostly tests. Databases carries the most marks on this paper.

Databases65 marks · 25%
Q1, Q2, Q3, Q4
Modelling and simulation65 marks · 25%
Q5, Q6, Q7, Q8
Web science65 marks · 25%
Q9, Q10, Q11, Q12, Q13
Object-oriented programming65 marks · 25%
Q14, Q15, Q16, Q17

Mentor insights

Traps to watch for

01
Databases: OnlyScience scientific journal database: ERD, SQL queries, views, · Q3
Clearly label both the relationship verbs ('publishes', 'responsible for') and the 1:N cardinalities.
02
Modelling and simulation: Natural language processing challenges and neural network charact · Q8
Structure your 6-mark contrast around the three linguistic levels: vocabulary (ambiguity), syntax (flexibility vs strictness), and semantics (context).
03
Web science: HTTP/HTTPS protocols, client-side JavaScript, and server-side scr · Q9
Describe the fundamental Request-Response cycle clearly between client and server.
04
Object-oriented programming: JETS LinkedList ADT operations for managing Specimen and unique S · Q17
Remember the two core features of an ADT: hidden internal representation and a standard set of operations.

Solutions and tips are GeorgeTheTutor's own; the questions are not reproduced here. Not affiliated with or endorsed by the International Baccalaureate Organization.

Want someone to go through your paper with you?

Our maths mentors took these exams recently. One-to-one, they'll find where your marks are going and fix it before May.

Find your maths mentor More past papers