NLP and the Four Main Categories of Ambiguity DATA-SCIENCE-OPTIMIZE Practice Questions (Page 4)

Questions 16–20

1. 16expert · hard

   A search engine must index and retrieve documents based on user queries. A user searches for 'jaguar speed.' The system must decide whether the user is interested in the animal or the car brand. Which type of ambiguity is this, and what is the most effective way to resolve it?

   • APragmatic ambiguity; use the user's location to determine intent
   • BSemantic ambiguity; use the user's search history and context to infer intent
   • CLexical ambiguity; use a thesaurus to expand the query with synonyms
   • DSyntactic ambiguity; use a parser to analyze the query structure
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   What is the key difference between semantic ambiguity and lexical ambiguity?

   • ASemantic ambiguity involves sentence structure, while lexical ambiguity involves word meanings
   • BSemantic ambiguity involves the combination of words, while lexical ambiguity involves a single word with multiple meanings
   • CSemantic ambiguity depends on speaker intent, while lexical ambiguity does not
   • DSemantic ambiguity is always resolved by context, while lexical ambiguity is not
   Reveal Solution Discussion Select an answer first
4. 18expert · hard

   A legal-tech company is building an NLP system to extract obligations from contracts. The system must handle both lexical ambiguity (e.g., 'consideration' meaning payment or thought) and syntactic ambiguity (e.g., 'the party shall notify the client of the breach in writing'). The team has limited labeled data and must choose a strategy. Which approach best addresses both types of ambiguity while minimizing manual annotation?

   • ATrain a traditional machine-learning model with hand-engineered features from scratch
   • BFine-tune a pre-trained transformer model on a small set of domain-specific examples
   • CUse a simple keyword-matching system that flags sentences containing ambiguous words
   • DUse a rule-based system with hand-crafted patterns for each ambiguous term and structure
   Reveal Solution Discussion Select an answer first
5. 19application · medium

   A news-aggregation system extracts articles and must classify them by topic. It encounters the headline: 'Bass wins award.' The system must decide whether 'bass' refers to a fish or a musical instrument. Which NLP task is specifically designed to resolve this type of ambiguity?

   • AWord-sense disambiguation
   • BDependency parsing
   • CPart-of-speech tagging
   • DNamed entity recognition
   Reveal Solution Discussion Select an answer first
6. 20expert · hard

   A healthcare NLP system must interpret clinical notes. It encounters the sentence: 'The patient was given a shot of penicillin.' The system must decide whether 'shot' refers to an injection or a small amount (e.g., a shot of medicine). Additionally, the sentence 'The doctor examined the patient with the stethoscope' must be interpreted. Which combination of ambiguity types is present in these two sentences?

   • ASyntactic in the first sentence and lexical in the second
   • BSemantic in both sentences
   • CLexical in the first sentence and syntactic in the second
   • DPragmatic in both sentences
   Reveal Solution Discussion Select an answer first