Assessment of Post-Project Paper 1

Dear NN,

We have finished assessing your post-project paper.

Your grade for this deliverable is: U

I have attached a more detailed assessment below.

Your paper was evaluated using the grading criteria specified on the course website (link). It was first assessed by one of the teaching assistants, and then an AI assistant provided a separate evaluation. I reviewed both of these assessments, resolved any differences between them, and combined them into a final evaluation.

I am sorry that I do not have better news for you. You can revise your paper based on the feedback in the assessment. The next opportunity for examination is in connection with the re-exam period in June. To participate in this examination, submit your revised paper through Lisam.

Please do not hesitate to reach out if you have any questions.

Best, Marco

Clarity

Your report presents an interesting project and includes useful diagrams and tables, but it lacks a clear and complete description of what the project set out to do. Important aspects of your experimental setup, such as model selection and class definitions, are either introduced too late or not explained thoroughly. For instance, it’s unclear which model produced the results in Section 1.5.2, and the precision values in Table 2 don’t align with earlier ones. You also use inconsistent terminology for your label classes, which creates confusion. The pipeline figure at the end of the Describe section is helpful but not well-integrated into the explanation. Finally, you incorrectly write “Near-entity models (NER)”; this should be “Named-Entity Recognition (NER)”.

Critical Reflection

Your reflection includes thoughtful observations about dataset inconsistencies and the challenges of annotation conversion, and you draw on relevant research to support your analysis. However, your report does not explicitly connect these reflections to specific ideas or skills from the course. This connection is essential. The course content—particularly on sequence labeling (Unit 4) and model evaluation—offers rich ground for reflection, but you need to make this link more explicit. In your next revision, take the opportunity to show how the course helped you frame or solve the challenges you encountered. You do not need to list lectures or labs, but you should reflect on how specific techniques, concepts, or skills learned in the course influenced your thinking or approach.

Articulation of Learning

You do a good job identifying meaningful lessons from your project—such as the importance of data quality, annotation consistency, and class balance—and you support your reflections with relevant research literature. You also mention areas for future improvement, including advanced rebalancing strategies and evaluation techniques. This shows that you have taken a broad, research-informed view of the problem. One area for improvement is how you evaluate and motivate your experiments. For instance, accuracy is not a suitable metric for this task, and the TAB paper you cited explicitly discusses why. This weakens your discussion and suggests a gap in your understanding of evaluation. In your revision, consider how your learning is reflected in your methodological choices, not just your conclusions.

Effort and Care

Your writing demonstrates care and professionalism overall. Terminology is generally used well, and your references are accurate and appropriately chosen. However, some grammatical errors—especially in the shared Describe section—have not been corrected, and there are a few issues with word choice (e.g., “autonomous anonymization”). These reduce the polish of your writing.

Overall Assessment

The overall assessment of the report is: Below expectation

While your report demonstrates engagement with relevant literature and shows that you learned important lessons during the project, it does not meet the minimum expectations in two critical areas: clarity and critical reflection. The report lacks a clear and complete account of your project setup and results, and it does not explicitly connect your reflections to the course content as required. You still have room in your report to address these issues. Strengthening the Describe section and making clear links between your project experience and the course material will significantly improve your report in the next revision.