Natural Language Processing (NLP) AIE Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   Which text preprocessing technique reduces words like 'running', 'runs', and 'ran' to a common base form such as 'run'?

   • AStemming or lemmatization
   • BStop-word removal
   • CTokenization
   • DPart-of-speech tagging
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A global brand wants to monitor social media sentiment across different regions. The same emoji can have different meanings in different cultures (e.g., the thumbs-up emoji is positive in some cultures but offensive in others). The team is considering: (1) using a global sentiment model trained on diverse data, (2) using a single sentiment model trained only on English data, (3) removing all emojis before analysis, (4) using a rule-based sentiment lexicon. Which approach is most likely to handle cultural nuances effectively?

   • AUse a global sentiment model trained on diverse multilingual and multicultural data.
   • BUse a rule-based sentiment lexicon.
   • CRemove all emojis before analysis.
   • DUse a single sentiment model trained only on English data.
   Reveal Solution Discussion Select an answer first
4. 23expert · hard

   A research team is building a text classification model for scientific papers. The papers contain many technical terms and acronyms. The team has a limited amount of labeled data. They are considering: (1) using a pre-trained language model with domain adaptation, (2) using a bag-of-words model with TF-IDF, (3) applying aggressive stemming, (4) using a word2vec model trained on the same corpus. Which approach is most likely to achieve the best classification performance?

   • AApply aggressive stemming to all words.
   • BUse a bag-of-words model with TF-IDF.
   • CUse a pre-trained language model with domain adaptation on scientific text.
   • DUse a word2vec model trained on the same corpus.
   Reveal Solution Discussion Select an answer first
5. 24application · easy

   A research team is analyzing a large dataset of product reviews. They want to reduce the dimensionality of the feature space while preserving the meaning of words like 'running' and 'ran'. Which preprocessing technique should they apply?

   • ALemmatization
   • BTokenization
   • CPart-of-speech tagging
   • DStop-word removal
   Reveal Solution Discussion Select an answer first
6. 25expert · hard

   A team is preprocessing a dataset of medical notes for a named entity recognition model. They must balance reducing vocabulary size with preserving clinical meaning. Which preprocessing choice is most appropriate?

   • AUse lemmatization instead of stemming to preserve the clinical meaning of words.
   • BUse aggressive stemming to maximize vocabulary reduction.
   • CRemove all stop words, including negation words like 'not'.
   • DSkip preprocessing entirely to preserve all original text.
   Reveal Solution Discussion Select an answer first