Phase 4 Exam Drill: System Design, Report Outlines & Ethics

NUST Model Answer [4 Marks]:
Establishing background context, formal Problem Statement, Research Aim (The "What"), Objectives (The "How"), Research Questions/Hypotheses, and Report Organization.

NUST Model Answer [4 Marks]:
Systematically reviewing existing research (SLR), PRISMA flow diagram, identifying knowledge gaps, and presenting theoretical frameworks.

NUST Model Answer [4 Marks]:
Detailing research paradigms, research methodology, software development methodology, sampling strategy, data collection instruments, and ethics.

NUST Model Answer [4 Marks]:
Presenting system requirements analysis, Context Diagrams, Use Case Diagrams, Activity Diagrams, 3-Tier System Architecture, ERD, and algorithm pseudocode.

NUST Model Answer [4 Marks]:
Documenting system implementation environment, empirical benchmarking results, statistical validation (t-test/ANOVA), error rate analysis, and conclusions.

Model Answer [3 Marks]:
Making up fake benchmark numbers, survey responses, or test logs.

Model Answer [3 Marks]:
Manipulating real benchmark logs or selectively deleting outliers to artificially fit a hypothesis.

Model Answer [3 Marks]:
Copying text, code, or ideas from other authors without proper attribution/citation.

Model Answer [3 Marks]:
Copying fake AI-generated paper citations or unverified code into a report without verification.

Model Answer [3 Marks]:
A high-level Data Flow Diagram (Level 0) representing the system as a single process interacting with external entities.

Model Answer [3 Marks]:
A UML diagram illustrating actors, system boundary, and functional use cases.

Model Answer [3 Marks]:
A UML diagram modeling sequential logic flows, decision nodes, and parallel activities.

Model Answer [3 Marks]:
A database design model showing entities, primary keys (PK), foreign keys (FK), and relationships (1-to-N).

Model Answer [3 Marks]:
1. Hardware / Perception Layer, 2. Application / Logic Layer, 3. Data Layer.

Model Answer [3 Marks]:
Providing kernel-space device access as an un-buffered stream of bytes for user-space applications.

Model Answer [3 Marks]:
alloc_chrdev_region() (or register_chrdev_region()).

Model Answer [3 Marks]:
1. Problem Statement, 2. Proposed Methodology.

Model Answer [2 Marks]:
Overleaf LaTeX editor.

Model Answer [3 Marks]:
A mandatory sub-task that must be executed whenever the base use case is triggered.

Model Answer [3 Marks]:
An optional execution branch triggered only under specific conditional logic.

Model Answer [3 Marks]:
Decomposing the Level 0 Context Diagram into main sub-processes and data stores.

Model Answer [3 Marks]:
A unique identifier column for records in a database table.

Model Answer [3 Marks]:
A column referencing the primary key of another table to establish relational constraints.

Model Answer [3 Marks]:
Systemic unfairness in model outputs caused by biased training data or flawed feature selection.

Model Answer [3 Marks]:
A model whose internal decision-making logic is opaque and non-interpretable to humans.

Model Answer [3 Marks]:
Initializes a `struct cdev` with file operations callback handlers.

Model Answer [3 Marks]:
A 200-300 word summary of problem, methodology, key findings, and conclusions.

Model Answer [3 Marks]:
Formally educating subjects on study risks/benefits and obtaining written permission before data gathering.

Model Answer [3 Marks]:
A premise accepted as true without direct proof to allow research to proceed (e.g. assuming active NFC cards).

NUST Model Answer [4 Marks]:
1. System Requirements Analysis (Functional & Non-Functional), 2. System Modeling (Context Diagram, Use Cases, Activity Diagrams), 3. System Architecture (3-Tier Decoupled Structure), 4. Database Design (ERD & Schema Dictionary).

Model Answer [5 Marks]:
Mapping function pointers: `.owner = THIS_MODULE`, `.read = driver_read`, `.write = driver_write`, `.open = driver_open`, `.release = driver_close`.

Model Answer [4 Marks]:
Layer 2 (Application) handles minutiae matching, isolating Layer 1 (Hardware sensors) from direct Layer 3 (SQLite) queries.

Model Answer [4 Marks]:
To prove to the ethics/proposal committee that the proposed topic addresses an un-solved literature gap.

Model Answer [4 Marks]:
Context Diagram shows external manual data flows; Use Case Diagram details internal functional interactions and actors.

Model Answer [4 Marks]:
Diamond decision nodes branch into `[Match == True]` (allow entry) or `[Match == False]` (deny entry & flag incident).

Model Answer [3 Marks]:
1. Anonymizing biometric minutiae templates, 2. Auditing false rejection rates across different demographic groups.

Model Answer [4 Marks]:
Poisons the scientific record with fake findings, leading other researchers to build on false premises.

Model Answer [4 Marks]:
Safely copying kernel-space buffer data into user-space application memory without causing kernel panic.

Model Answer [4 Marks]:
1. Implementation environment, 2. Code integration, 3. Empirical results & graphs, 4. Conclusions & future work.

Model Answer [3 Marks]:
To enforce referential integrity between entities (e.g. Student and Attendance tables).

Model Answer [4 Marks]:
LLMs can generate fake author names, fabricated volume numbers, and non-existent paper titles that fail audit checks.

Model Answer [3 Marks]:
Line graph showing Memory Overhead (MB) on Y-axis vs Payload Size (KB) on X-axis.

Model Answer [3 Marks]:
Providing a single-paragraph summary of topic, problem statement, aim, methodology, and expected contribution.

Model Answer [4 Marks]:
Functional: What system does (e.g. scan fingerprint).
Non-Functional: System quality attributes (e.g. latency < 3.5s, 99.9% uptime).

Model Answer [4 Marks]:
Safely copying user-space write payload buffer into kernel-space device memory.

Model Answer [3 Marks]:
Automated BibTeX citation management, superior mathematical formatting, and consistent IEEE/ACM layouts.

Model Answer [3 Marks]:
Stating foundational preconditions assumed true so the study scope remains tractable.

Model Answer [3 Marks]:
1. Secure local DB encryption, 2. Restricting access to authorized invigilators only.

Model Answer [4 Marks]:
To formally specify algorithmic logic in a language-agnostic format for peer verification.

Model Answer [3 Marks]:
Using synchronization fork and join bars.

Model Answer [4 Marks]:
Because automated tools (Turnitin) flag text matching, whereas detecting falsified logs requires raw data audits.

Model Answer [4 Marks]:
Justifying the research design, sampling methods, data collection tools, and ethical frameworks.

Model Answer [3 Marks]:
Logging kernel-space diagnostic messages to `dmesg` buffer.

Model Answer [3 Marks]:
1. `STUDENT` (student_id, full_name, nfc_uid, minutiae_hash), 2. `ATTENDANCE` (record_id, student_id, scan_time, status).

Model Answer [4 Marks]:
Chapter 5 results are compared directly against the literature baselines documented in Chapter 2.

Model Answer [3 Marks]:
Releasing major/minor device numbers back to kernel memory upon driver module cleanup.

Model Answer [3 Marks]:
To prove project feasibility within the academic semester timeframe.

Model Answer [4 Marks]:
Data Layer (Tier 3) is isolated behind Tier 2 APIs, preventing direct unauthorized hardware access to database files.

Model Answer [4 Marks]:
Alters technical accuracy, introduces awkward phrasing, and fails Turnitin AI-detection checks.

NUST Model Answer [20 Marks Breakdown]:
1. Ch 1: Introduction [4M]: Background, Problem Statement, Aim, Objectives (3-4), RQs/Hypotheses, Significance, Report Structure.
2. Ch 2: Literature Review [4M]: SLR methodology, PRISMA flow diagram, State-of-the-art baselines, Knowledge gaps, Theoretical framework.
3. Ch 3: Methodology [4M]: Research paradigm (Positivism/Pragmatism), Applied Research design, Software Dev methodology (Iterative Prototyping), Sampling, Data instruments, Ethics.
4. Ch 4: Analysis & Design [4M]: Functional/Non-functional requirements, Context Diagram, Use Case Diagrams, Activity Diagrams, 3-Tier Architecture, ERD Schema.
5. Ch 5: Implementation & Testing [4M]: Code environment, Driver integration, Empirical benchmarking, Statistical validation (t-test/ANOVA), Conclusions & Future work.

NUST Model Answer [20 Marks Breakdown]:
1. Abstract & Background [4M]: Kernel module execution in openEuler, character vs block devices.
2. Setup & Controls [5M]: QEMU ARM64 setup, openEuler kernel 5.10, memory isolation, gcc compilation flags.
3. Driver Architecture & Methods [5M]: `alloc_chrdev_region()`, `cdev_init()`, `file_operations` struct (`read`, `write`, `ioctl`), `copy_to_user()` memory safety.
4. Empirical Results & Discussion [6M]: Benchmarking throughput (MB/s) and latency ($\mu s$) across payload sizes (64B-64KB) with line graphs and statistical confidence analysis.

Model Answer [20 Marks]:
1. Title & Intro [3M], 2. Background [3M], 3. Problem Statement & Aim [3M], 4. Objectives & SLR/PRISMA [4M], 5. Methodology & System Design (Context/Use Case/3-Tier) [4M], 6. Assumptions & Harvard References <5 yrs [3M].

Model Answer [5 Marks]:
Chapter 3 specifies the software methodology rules (e.g. Iterative Prototyping), while Chapter 4 details the exact architectural blueprints (Use Cases, ERD, 3-Tier) built during those sprint iterations.

Model Answer [4 Marks]:
Flawed proposal: Unstated assumptions leave room for hidden scope failure (e.g. assuming pre-existing student NFC cards).

Model Answer [5 Marks]:
Unchecked memory pointer copying causes invalid kernel memory dereferences, triggering kernel panics and corrupting test log reliability.

Model Answer [5 Marks]:
Functional requirements define *what* the system must achieve; architecture defines *how* technical layers decouple components to fulfill those goals securely.

Model Answer [5 Marks]:
1. No fabricated data, 2. No falsified/cherry-picked runs, 3. Turnitin-verified citations, 4. Verified non-hallucinated AI references, 5. IRB-cleared human data.

Model Answer [4 Marks]:
Proves relational database normalization, eliminating data redundancy and ensuring atomic attendance audit logging.

Model Answer [4 Marks]:
Context Diagram shows whole system boundary as 1 process; DFD Level 1 explodes it into core sub-processes and data stores.

Model Answer [5 Marks]:
Violates ethical data privacy standards; unencrypted minutiae hashes risk extraction and identity impersonation breaches.

Model Answer [4 Marks]:
Using decision nodes leading to fallback error branches (e.g. routing to invigilator manual desk after 3 failed scans).

Model Answer [4 Marks]:
Tables provide exact numeric values for replication; line graphs visually illustrate performance scaling trends across payload sizes.

Model Answer [4 Marks]:
Flawed title: Vague, unscientific, and lacks technical specificity. Corrected: "A Dual-Factor Biometric Examination Identity Authentication System for Higher Education."

Model Answer [4 Marks]:
BibTeX automatically formats in-text citations and reference lists into strict Harvard/IEEE rules directly from raw DOI metadata.

Model Answer [5 Marks]:
Kernel log timestamps recorded via `dmesg` provide verifiable raw data for computing driver throughput and context switch overhead.

Model Answer [4 Marks]:
`«include»`: Mandatory sub-use case executed every time.
`«extend»`: Optional sub-use case executed only under specific conditional triggers.

Model Answer [4 Marks]:
To ensure Chapter 4 system design artifacts are engineered specifically to test and answer those initial research questions.

Model Answer [5 Marks]:
Storing raw images is a major privacy violation (reconstructible fingerprint data); storing encrypted minutiae hashes protects biometric privacy.

Model Answer [4 Marks]:
By succinctly demonstrating a clear problem gap, rigorous methodology, feasible system design, and adherence to ethics rules.

NUST Model Answer [5 Marks]:
ABSOLUTELY NOT! Examiner Trap: Regardless of software development methodology (Agile, Waterfall, Prototyping), NUST CS final year reports explicitly mandate Chapter 4 (Analysis & Design) to detail Context Diagrams, Use Cases, ERD, and 3-Tier Architecture blueprints.

Model Answer [5 Marks]:
YES! Code plagiarism is equal to text plagiarism. All external open-source driver routines must be explicitly cited in comments and text references.

Model Answer [4 Marks]:
NO! Examiner Flaw: Use Case Diagrams model interactions between actors and use cases inside the system boundary. Data flows belong in DFDs.

Model Answer [5 Marks]:
NO! Even with 12% overall similarity, if a single paragraph or code block is copied verbatim without citation, it constitutes plagiarism.

Model Answer [4 Marks]:
NO! Char drivers execute in kernel space (Ring 0) and must be compiled as kernel modules (`.ko`) and loaded via `insmod`.

Model Answer [4 Marks]:
YES! Any research collecting data from human participants (regardless of sample size or friendship) requires formal ethical clearance.

Model Answer [4 Marks]:
Clutters the report body. Chapter 5 should present summary tables/graphs; raw 10,000-line logs belong in the Appendix.

Model Answer [4 Marks]:
NO! Relational database rules strictly mandate that Primary Keys must be non-null unique identifiers.

Model Answer [5 Marks]:
NO! Generating fake performance benchmark data via AI constitutes Data Fabrication and voids empirical validity.

Model Answer [5 Marks]:
Presenting only screenshots of software user interfaces without providing empirical benchmark tables, statistical tests, or error rate analyses.