Natural Language Processing Week 10 NPTEL Assignment Answers 2025

NPTEL Natural Language Processing Week 10 Assignment Answers 2024

1. Common Steps of Entity Linking are –

Options:

• A) Reference Disambiguation → Candidate Selection
• B) Reference Disambiguation → Candidate Selection → Mention Identify
• C) Mention Identify → Candidate Selection → Reference Disambiguation
• D) All of the above

✅ Answer: C
Explanation:
Entity Linking follows this correct order: Mention Identification → Candidate Selection → Reference Disambiguation.

---

2.

✅ Answer: B

---

3. What is the commonness of (s, c₂) in the above question?

Options:

• A) 0.765
• B) 0.389
• C) 0.453
• D) 0.910

✅ Answer: c
Explanation:
Commonness typically refers to the probability that a surface form s refers to candidate entity c₂.

---

4. Higher value of keyphraseness represents higher probability of:

Options:

• A) An article to get selected as linkable candidate
• B) A phrase to get detected as a mention
• C) An article to get disambiguated from other candidates
• D) None of the above

✅ Answer: b
Explanation:
Keyphraseness measures how likely a phrase is a named entity mention.

---

5. Which of the following problems exist in bootstrapping technique for Information Extraction?

Options:

• A) Sensitiveness towards the seed set
• B) High precision
• C) Less manual intervention
• D) All of the above

✅ Answer: a
Explanation:
Bootstrapping heavily relies on the quality and bias of the initial seed set, making it sensitive to errors.

---

6. Which of the following is an advantage of unsupervised relation extraction?

Options:

• A) Can work efficiently with small amount of hand-labeled data
• B) Not easily generalizable to different relations
• C) Need training data
• D) Always perform better than supervised techniques
• E) None of the above

✅ Answer: e
Explanation:
None of the above are advantages. Unsupervised relation extraction is hard to evaluate, often less accurate.

---

7. Which of the following is/are a Hearst’s Lexico-Syntactic Pattern for automatic acquisition of hyponyms?

Options:

• A) X or other Y
• B) X and other Y
• C) Y including X
• D) X but not Y

✅ Answer: a, b, c
Explanation:
Hearst patterns help extract hyponym-hypernym relationships. Examples include:

• “such as”,
• “including”,
• “and other”,
• “or other”.

---

8. Advantage of Distant Supervision over Bootstrapping method

Options:

• A) Need more data
• B) Less human effort
• C) Can handle noisy data better
• D) No Advantage

✅ Answer: c
Explanation:
Distant Supervision is more robust in handling noise, especially when labeled data is noisy or incomplete.

---

9. For a dataset with few but important relations, the most useful metric is:

Options:

• A) Precision
• B) Recall
• C) Accuracy
• D) F1-Score

✅ Answer: b
Explanation:
In scenarios where missing important relations is critical, Recall is prioritized to capture as many as possible.

---

10. Bootstrapping can be considered as:

Options:

• A) Supervised Approach
• B) Unsupervised Approach
• C) Semi-supervised Approach
• D) All of the above
• E) None of the above

✅ Answer: c
Explanation:
Bootstrapping starts with a small labeled set (supervised) and grows using unlabeled data, making it semi-supervised.