Project Assessment Criteria

English עברית

Student project assessment criteria. The five dimensions a project is judged on, what each one expects in practice, and the anti-patterns that most often weaken it. For the detailed, requirement-by-requirement checklist, see Project Requirements.

Criterion 1

Demonstrate Problem-Domain Understanding

Students should develop substantial expertise in the project domain. The analysis should go beyond common-sense observations and be supported by research, professional sources, and relevant literature.

Students should clearly understand
  • The practical use case and the problem being addressed.
  • The value of solving the problem.
  • The relevant users, stakeholders, and operational constraints.
  • Existing solutions, methods, and limitations.
Anti-patterns
  • Superficial understanding of the problem.
  • Motivation based on generic AI trends rather than a real user need.
  • Little or no discussion of existing solutions.
  • Incorrect or inconsistent use of domain terminology.
  • Project objectives that are vague or disconnected from the actual problem.
Criterion 2

Explore the Solution Space in Depth

Students should thoroughly investigate the possible solution space rather than immediately selecting the first feasible approach.

The project should demonstrate
  • Awareness of state-of-the-art models, methods, and technologies.
  • Comparison of alternative solution approaches.
  • Clear mapping between problem-domain requirements and technical design decisions.
  • Well-reasoned selection of models, architectures, tools, and system components.
  • Understanding of the advantages, limitations, and trade-offs of the selected solution.
Anti-patterns
  • Choosing the first model or framework without justification.
  • Technology-driven rather than problem-driven design.
  • No comparison with alternative approaches.
  • Design decisions based on popularity or familiarity instead of requirements.
  • Inconsistent mapping between problem characteristics and technical choices.
Criterion 3

Ensure Novelty and Effective Reuse

The project should contain a meaningful element of novelty in the problem formulation, solution design, system integration, experimentation, or application. At the same time, students should effectively reuse existing models, libraries, datasets, tools, and software components where appropriate, and should not unnecessarily recreate functionality that is already available.

Students should clearly distinguish between
  • Existing components that were reused.
  • Components that were adapted or extended.
  • New contributions developed as part of the project.
Anti-patterns
  • Reimplementing standard libraries or models without a clear reason.
  • Building large amounts of infrastructure instead of solving the research problem.
  • Copying existing projects with only cosmetic modifications.
  • Claiming novelty where only implementation details differ.
  • Failing to identify what is genuinely new in the project.
Criterion 4

Apply a Correct and Rigorous Methodology

Students should use a methodology appropriate to the project and problem domain. The evaluation should demonstrate whether the proposed solution actually addresses the original problem and provides measurable value.

This includes
  • Correct data collection, selection, generation, or preprocessing procedures.
  • Appropriate separation of training, validation, and test data.
  • Selection of suitable baselines.
  • Use of relevant evaluation metrics.
  • Proper experimental design.
  • Analysis of errors, limitations, and failure cases.
  • Reproducible and well-documented experiments.
Anti-patterns
  • Using inappropriate datasets or evaluation metrics.
  • Data leakage between training and testing.
  • Comparing against weak or missing baselines.
  • Reporting only successful examples while ignoring failures.
  • Drawing conclusions unsupported by experimental evidence.
  • Irreproducible experiments or undocumented procedures.
Criterion 5

Present and Defend the Project Clearly

Students should present the problem, methodology, solution, and results clearly and professionally. Explanations should be short, precise, and informative rather than vague, overly general, or unnecessarily detailed.

Students should
  • Use accurate technical and domain-specific terminology.
  • Explain the project concisely and without irrelevant information.
  • Justify all important design and implementation decisions.
  • Explain the internal operation of the models and system components.
  • Present results using suitable tables, figures, examples, and comparisons.
  • Clearly distinguish between assumptions, evidence, conclusions, and limitations.
  • Answer questions confidently and demonstrate personal understanding of the project.
Anti-patterns
  • Slides overloaded with text or irrelevant details.
  • Buzzwords without technical understanding.
  • Inability to explain design decisions or model behavior.
  • Confusing implementation details with project contributions.
  • Weak visual presentation of results.
  • Long, unfocused answers that fail to address the question directly.

Quick Links