🚨 Quiz Instructions:
Attempt all questions first.
✔️ Click SUBMIT at the end to unlock VIEW ANSWER buttons.
Quiz Mode:

Word Sense Disambiguation (WSD) MCQs with Answers and Explanations

Understanding the correct meaning of a word in context is one of the most fundamental challenges in Natural Language Processing (NLP). This collection of multiple-choice questions (MCQs) focuses on Word Sense Disambiguation (WSD), covering core concepts, algorithms, and real-world scenarios. These questions are designed for students, researchers, and professionals preparing for exams, interviews, or strengthening their NLP knowledge.

What is Word Sense Disambiguation (WSD)?

Word Sense Disambiguation (WSD) is the task of identifying the correct meaning (sense) of a word based on its context in a sentence or document. Many words in natural language are ambiguous (polysemous), meaning they have multiple meanings.

For example:

  • "bank" → Financial institution OR river edge
  • "bat" → Flying mammal OR sports equipment
  • "plant" → Industrial facility OR living organism

WSD systems analyze surrounding words, syntax, semantics, and sometimes external knowledge bases to determine the intended meaning.

Why is WSD Important?

  • Improves machine translation accuracy
  • Enhances search engines and information retrieval
  • Supports chatbots and virtual assistants
  • Essential for text understanding and AI applications

Common Approaches to WSD

  • Knowledge-based methods (e.g., Lesk algorithm, WordNet)
  • Supervised learning (trained on labeled datasets)
  • Unsupervised methods (Word Sense Induction)
  • Neural approaches (contextual embeddings like transformers)

What You Will Learn from These MCQs

  • Context-based ambiguity resolution
  • Lesk algorithm and gloss overlap
  • Most Frequent Sense (MFS) bias
  • Graph-based and probabilistic WSD
  • Transformer-based contextual understanding
  • Domain adaptation and sense granularity issues

Each question includes a detailed explanation to help you clearly understand why a particular sense is correct, making this set ideal for both learning and revision.

📚 Table of Contents

1.
A WSD system attempts to disambiguate the word "bank" in the sentence: "The fisherman sat on the bank and watched the river flow." The algorithm only considers a context window of one word on each side. What is the most likely issue?






Correct Answer: B

Explanation:

A very small context window may not include important words like “river” that indicate the intended meaning. Without enough surrounding context, the WSD algorithm cannot gather sufficient evidence to correctly determine the sense of the word.

2.
A simplified Lesk algorithm compares gloss overlaps for the word "bass". Gloss 1: "a type of fish found in rivers". Gloss 2: "a low-frequency musical tone". Sentence: "The bass was hard to hear during the concert." Which sense will the algorithm most likely choose?






Correct Answer: C

Explanation:

The Lesk algorithm chooses the sense whose dictionary gloss has the largest overlap with the context words. In this sentence, words like “hear” and “concert” relate to music, making the musical sense the most appropriate interpretation.

3.
A supervised WSD model always predicts the "financial institution" sense for the word "bank", even when used in river-related contexts. What problem does this illustrate?






Correct Answer: A

Explanation:

Many WSD systems learn to predict the sense that appears most frequently in training data. This leads to bias toward the Most Frequent Sense (MFS), causing incorrect predictions when rarer senses appear in different contexts.

4.
In a graph-based WSD system, nodes represent WordNet synsets and edges represent semantic relations. The algorithm selects the sense with the highest PageRank score. What intuition does this reflect?






Correct Answer: D

Explanation:

Graph-based WSD assumes that the correct sense will have strong semantic connections with other senses appearing in the same context. PageRank helps identify the sense that is most central or influential within this semantic network.

5.
Consider the paragraph: "She deposited money in the bank yesterday. Later she returned to the bank to withdraw cash." A WSD system assigns the same sense of "bank" in both sentences. Which assumption supports this behavior?






Correct Answer: B

Explanation:

The Yarowsky algorithm relies on the principle "one sense per discourse," which suggests that a word appearing multiple times in the same document usually carries the same meaning throughout that discourse.

6.
A system groups occurrences of the word "crane" into clusters based on contexts like {bird, fly, nest} and {construction, steel, lift}. What task is this system performing?






Correct Answer: C

Explanation:

Word Sense Induction (WSI) automatically discovers senses by clustering contexts in which a word appears. Unlike traditional WSD, it does not rely on predefined sense inventories such as WordNet.

7.
A transformer model processes the sentence: "The mouse clicked on the icon." Why can it correctly interpret "mouse" as a computer device rather than an animal?






Correct Answer: A

Explanation:

Transformer models generate contextual embeddings where the representation of a word changes depending on surrounding words. In this case, the presence of "clicked" and "icon" signals the computer-related sense.

8.
A WSD system trained on news articles performs poorly on biomedical text. What is the most likely reason?






Correct Answer: D

Explanation:

Different domains often use words with different senses or frequency distributions. A model trained on news data may not recognize specialized meanings commonly used in biomedical literature.

9.
A neural WSD model compares the embedding of a sentence context with embeddings of candidate gloss definitions. Which principle does this approach rely on?






Correct Answer: C

Explanation:

Gloss-based neural WSD methods encode both the sentence context and dictionary glosses into vector representations. The sense whose gloss embedding is most similar to the context embedding is selected.

10.
Sentence: "The plant produces electricity using turbines." Which sense of the word "plant" is most appropriate?






Correct Answer: A

Explanation:

Words like "produces" and "turbines" strongly indicate an industrial context. Therefore, the correct interpretation of "plant" here refers to a power or industrial facility rather than vegetation.

11.
Sentence: "He went to the bank." Why is this sentence difficult for a WSD system to disambiguate?






Correct Answer: C

Explanation:

The sentence provides no contextual indicators to determine whether "bank" refers to a financial institution or a river bank. Without surrounding words that provide semantic clues, WSD algorithms struggle to determine the intended sense.

12.
Sentence: "The python was spotted near the forest." Considering the ecological context, which sense of "python" is most likely?






Correct Answer: B

Explanation:

Context words such as "spotted" and "forest" indicate a wildlife environment. These contextual cues strongly suggest that the word "python" refers to the snake rather than the programming language.

13.
Two sentences are given: 1. "She opened the book." 2. "The police will book the suspect." Why can contextual embedding models correctly distinguish the meaning of "book" in these sentences?






Correct Answer: D

Explanation:

Contextual language models generate different vector representations for the same word depending on surrounding words. Thus, "book" in a reading context and "book" in a legal context produce different embeddings.

14.
A WSD system correctly predicts 70 senses out of 100 test instances. What is the accuracy of the system?






Correct Answer: A

Explanation:

Accuracy is calculated as the number of correct predictions divided by the total number of instances. Here it is 70 / 100 = 0.70, meaning the system correctly disambiguated 70% of the test cases.

15.
Sentence: "He fixed the bug in the code." Which feature is most useful for correctly identifying the sense of "bug" in this context?






Correct Answer: C

Explanation:

The word "code" strongly suggests a programming environment. Collocations with technical terms help WSD systems infer that "bug" refers to a software error rather than an insect.

16.
A WSD system struggles because WordNet lists many fine-grained senses for a word, but the dataset uses only a few broad meanings. What problem does this illustrate?






Correct Answer: B

Explanation:

Sense granularity refers to the mismatch between very fine-grained lexical senses in resources like WordNet and the broader senses actually used in real datasets or applications.

17.
Sentence: "He kicked the bucket yesterday." Why might a traditional WSD system fail to interpret this correctly?






Correct Answer: D

Explanation:

The phrase "kick the bucket" is an idiom meaning "to die." If the system analyzes each word independently instead of recognizing the phrase as a multiword expression, it may incorrectly interpret the literal meaning.

18.
A probabilistic WSD model estimates the following probabilities: P(financial bank | context) = 0.3 P(river bank | context) = 0.7 Which sense should the model choose?






Correct Answer: A

Explanation:

In probabilistic WSD models, the sense with the highest posterior probability given the context is selected. Since 0.7 is greater than 0.3, the river bank sense is chosen.

19.
A transformer model assigns the highest attention weight to the word "river" when predicting the sense of "bank". What does this indicate?






Correct Answer: C

Explanation:

Attention mechanisms allow transformer models to identify and emphasize the most informative context words. In this case, "river" provides strong evidence that "bank" refers to the river edge sense.

20.
A multilingual WSD model uses translation alignment to disambiguate the word "bat" in the sentence: "The bat flew out of the cave." The aligned Spanish translation is "murciélago". Why does this help?






Correct Answer: B

Explanation:

Cross-lingual WSD relies on translations to resolve ambiguity. The Spanish word "murciélago" specifically refers to the flying mammal, eliminating the alternative sense of "bat" as a sports equipment.