Natural Language Processing
CSE · Midterm · Spring 24
Batch 58
Read the questions as text
Read off the PDF by AI, so it may have mistakes. The PDF is the original.
[Figure: University Logo]
Daffodil International University
Department of Computer Science and Engineering
Faculty of Science & Information Technology
Midterm Examination, Spring-2024
Course Code: CSE445, Course Title: Natural Language Processing
Level: 4 Term: 1 Batch: 58
Time: 1.5 Hours
Marks: 25
Answer ALL Questions
[The figures in the right margin indicate the full marks and corresponding course outcomes. All portions of each question must be answered sequentially.]
1. Assume the given corpus:
Walking is a good exercise. Mr. M. A. Sattar walks every morning. His email is sattar_143@diu.edu.bd
a) Apply the following preprocessing techniques on the corpus and determine the final outcome.
i) Remove punctuation and special character
ii) Tokenization
iii) Remove stopwords (determiner, auxiliary verbs)
iv) Lemmatization
[3] CO1
b) Construct a Regular Expression that can identify any email (email consists of any small letter, number and special character like "_", "." and "@").
[2]
2. Apply the Hidden Markov model to find out the Emission and Transition Probabilities considering the corpus below:
1. Sandy <Noun> Maya <Noun> enjoy <Verb> university <Noun>.
2. Can <Modal> Sandy <Noun> find <Verb> university <Noun>?
3. Will <Modal> Maya <Noun> find <Verb> university <Noun>?
4. Sandy <Noun> can <Modal> observe <Verb> Will <Noun>.
5. Maya <Noun> will <Modal> cherish <Verb> MIT <Noun>.
6. Will <Noun> is <Verb> student <Noun>.
[4] CO1
b) Determine the appropriate sequence of tags for the sentence, "Will can find campus." from the above probabilities.
[3]
3. a) Explain whether the following Context free grammar G is suitable for the CYK algorithm mentioning appropriate reason:
[2] CO2
S -> AC | BD
A -> AA | a
B -> CC | b
C -> AB | c
D -> DC | d
b) If possible, apply the CYK algorithm step by step using the formula to determine if "cadbba" is in L(G).
[6]
4. Consider the following documents:
Doc1: <Bangladesh is an independent country.>
Doc2: <Bangladesh became independent with a liberation war.>
Doc3: <Myanmar become independent this year.>
a) Build the vocabulary of the corpus (alphabetic order) and vectors for the documents (based on the ordered vocabulary) after applying the following preprocessing:
i) Remove punctuation and special character
ii) Convert the corpus into small letter
iii) Tokenization
iv) Remove stopwords (determinants and prepositions)
[2] CO2
b) Determine the similarity between Doc2 and Doc3 using cosine similarity and Euclidean distance.
[3]