Mark two location mentions using the exact character-offset contract. How would you check your result?

Instruction: Use the supplied source and offset contract. Return both LOCATION spans, then explain how you would check their boundaries without changing the text.

Context:

Practice precise entity labeling using an explicit offset convention. Explain how you would catch a boundary error before submitting the spans.

Fictional practice task

Exact source text:

Ship from York to New York.

Annotate both complete location mentions as LOCATION. Offsets count Unicode code points from zero in this exact string. Use half-open intervals [start, end), so the start is included and the end is excluded. Spaces and punctuation count as characters. Do not trim, normalize or otherwise change the source. Return each mention with its start and end, then explain how you would verify the spans. This is an exercise-specific offset contract.

Updated

Example Answer

I’d label York as LOCATION with start 10 and end 14, and New York as LOCATION with start 18 and end 26. The end is exclusive, so neither span includes the following space or final period. I’d treat New York as one complete mention rather than adding a third annotation for the York inside it.

I’d check both spans against the unchanged source: slicing from 10 to 14 must return York, and slicing from 18 to 26 must return New York. I’d also check the annotation preview to make sure the highlights cover exactly those words. This task counts Unicode code points, so I wouldn’t substitute a byte count or a different tool’s offset convention. If the export and preview disagreed, I’d resolve that before submitting.

Checked result

  • York: LOCATION, start 10, exclusive end 14. Check: source[10:14] returns York.
  • New York: LOCATION, start 18, exclusive end 26. Check: source[18:26] returns New York.

Make it your own

If you have used an annotation tool, name the check you actually performed: inspecting highlighted spans, comparing an export with the source, or validating the tool’s offset units. If this is your first exercise, explain these checks as a proposed method rather than claiming prior project experience.

Why this works

The answer supplies the actual labels and boundaries, then checks them against the source. It distinguishes the full New York mention from its embedded word and follows this exercise’s indexing contract rather than an assumed default.

Interviewer follow-up

What would you do if a highlighted span looked correct but the exported offsets were different?

I’d compare the original source, the tool’s stored text and the export’s documented offset convention. The display could look right while using different indexing units or an altered copy of the text. I’d reproduce the mismatch with an approved example and report the tool or contract issue through the project channel. I wouldn’t manually shift every offset until I understood what caused the difference.

Assessment criteria

  • Strong: Returns both correct LOCATION spans; verifies exact source slices and exclusive ends; avoids an extra nested mention; explains how to check the tool’s units without changing the source.
  • Adequate: Returns the two correct spans and explains at least one reliable boundary check, with no source edits or invented third location.
  • Weak: Counts words instead of characters, includes adjacent punctuation, labels only the inner York in New York, or changes the text before counting.

A tempting weak answer

“I’d highlight the two names and let the tool work out the offsets.”

A tool may help, but this answer never verifies the exported contract. A correct-looking highlight does not establish that the stored offsets use the required units or boundaries.

References

The practice text and offset convention are fictional. The public role description illustrates text annotation work; it does not prescribe this exercise’s convention.

Your preparation path

Choose the track that matches the role. Work through its questions in order, then explain each answer in your own words.

1. Entry level annotation

Apply guidelines, label text and spans, and explain a small practice project.

  1. How would you explain the data annotator role and the kind of work you would expect to do? Free sample
  2. How would you learn a new annotation guideline before starting your first batch? Member answer
  3. Apply a sentiment guideline to four short comments. Which labels would you choose, and why? Free sample
  4. Mark two location mentions using the exact character-offset contract. How would you check your result? Free sample
  5. Walk me through an annotation or quality-checking project you can discuss, including your own contribution and limits. Member answer

2. AI response evaluation

Compare responses using separate criteria for correctness, instruction following, and writing quality.

  1. Compare two AI responses against a supplied fact sheet. Which response is better under the rubric? Free sample
  2. One response is accurate but breaks the required format; another follows the format but contains a false claim. How would you rate them? Member answer
  3. Write a short rating rationale that identifies the decisive error without restating both responses. Member answer
  4. An AI response includes a factual claim you cannot verify from the supplied sources. What would you do? Member answer
  5. Two responses have different strengths and neither clearly wins. How would you apply the ranking rules? Member answer

3. Senior review and quality

Work through disagreement, missed critical cases, changing guidelines, review capacity, and reviewer calibration.

  1. Two experienced reviewers disagree repeatedly, and the deadline leaves little time for adjudication. What would you recommend? Free sample
  2. A batch has 98% accuracy against reviewed references but misses every critical item. Would you accept it? Member answer
  3. A labeling rule changes halfway through a delivery. Would you relabel old work, split the dataset or delay the release? Member answer
  4. The delivery requires review of every item, but the available reviewers cannot finish by the deadline. What would you change? Member answer
  5. A reference answer appears to contradict the written rule, and workers are being penalized for disagreeing with it. What would you do? Member answer

Try the 20 minute mock assessment. Use the fictional cases to practice; the self-check is not an employer's hiring benchmark.

Related Questions