NLP Lecture II

What is a binary bag of words?

• A) A bag used to store words
• B) A technique for text vectorization
• C) A type of vocabulary list
• D) A specific type of document

Which vectorization technique takes word importance and frequency into account?

• A) Bag of Words (BoW)
• B) Binary Bag of Words (BBoW)
• C) TF-IDF
• D) N-grams

What is the main advantage of using TF-IDF over simple BoW?

• A) Simplicity
• B) Higher dimensionality
• C) Word importance and frequency consideration
• D) Larger corpus size

In the context of text vectorization, when might using n-grams be more advantageous than simple word tokenization?

• A) When dealing with small text corpora
• B) When you need to preserve the order of words
• C) N-grams are never more advantageous
• D) When working with images, not text

Consider a document-term matrix for text vectorization, where rows represent documents and columns represent terms (words). How could we extract feature vectors for each word ?

• rows can be used as feature vectors

• columns can be used as feature vectors

• We need first to convert the matrix to a BoW matrix

• use countvectorizer(Binary=True)

Suppose you have a text corpus with hundreds of thousands of documents. You're using TF-IDF for vectorization. What is the potential issue you might encounter with such a large corpus when computing the TF-IDF matrix?

• A) The matrix will be too small to handle efficiently
• B) The dimensionality of the matrix becomes very high
• C) TF-IDF is not suitable for large corpora
• D) The computation time is reduced

What is the primary objective of Named Entity Recognition (NER) in natural language processing?

• A) Identifying and classifying specific entities in text, such as names of people, places, and organizations.
• B) Analyzing sentence structure and grammar to determine overall text sentiment.
• C)Identifying Names of persons in the documents

• D) Counting the frequency of stop words in a text corpus.

In the context of natural language processing, what does "predictive power" of a word refer to?

• A) The ability of a model to generate accurate predictions based on historical data.
• B) The capability of a word to predict future events in a text.
• C) The significance of a word's presence in a document based on its frequency and distribution.
• D) The degree of confidence in the correctness of the word syntax

What does a "vector space model" (VSM) represent in natural language processing?

• A) A model that predicts the future occurrences of words in a text corpus.
• B) A model for encoding and representing documents and or words as fixed-length numeric vectors.

• C) A model used for converting text to vectors of different lengths .

• D) A model that encodes documents only as fixed length numeric vectors